Content uploaded by Helen Perks
Author content
All content in this area was uploaded by Helen Perks on Mar 08, 2018
Content may be subject to copyright.
Global network configuration for
innovation: a study of international
fibre innovation
Helen Perks
1
and Richard Jeffery
2
1
Manchester Business School, University of Manchester, Booth St. West, Manchester, M15 6PB,
UK. h.perks@manchester.ac.uk
2
Northwest Textile Network, PO Box 295, Newtown Mill, Lees Street, Pendlebury, Swinton,
Manchester, M27 6WS, UK. r.jeffery@nwtexnet.co.uk
This paper develops and empirically investigates the notion of network configuration for
innovation. In many industries, firms are increasingly locked into a state of network
innovation. Innovation, in such contexts, is often driven by those firms who configure the
network to access and control critical innovation knowledge widely dispersed throughout the
network. The paper presents the findings of an in-depth study of the evolution of three
innovation networks in the global fibre industry. The mechanisms by which firms configure
extended networks throughout the innovation processes are unravelled and discussed. A
typology of three configuration types derived from the finding is forwarded. The findings
suggest that successful innovation network configuration involves recognising where the
innovation value resides in the network and developing capabilities and mechanisms to
understand and access such value. However, this is problematic for firms embedded in their
own base of knowledge and patterns of relationships. Specific managerial implications and
suggestions for future research are forwarded.
1. Introduction
In many industries, the source and enactment of
innovation has shifted from the single firm to a
group of networked firms. Increasingly firms seek
beyond their organisational boundaries to
integrate resources and capabilities core to their
ability to develop innovative products and
services. Critical innovation resources frequently
reside in a network and not in the firm alone
(Afuah, 2000). Consequently, many industries
now display features of high levels of innovation
inter-dependency, where development activities
by single firms have repercussions across the
extended industry network. However, the state
of understanding of how firms configure and
manage an extended network for innovation
is weak.
This study offers insight into the dynamics of
global network configuration for innovation. By
mapping the development of innovation networks
across a global industry (fibre industry), it un-
ravels the mechanisms by which firms configure
extended networks throughout the innovation
process. It further offers explanations behind the
observed firm behaviour and draws out implica-
tion for innovation management.
The first part of the paper reviews previous
research in the networks and innovation domains
and develops the notion of network configuration
for innovation. It suggests that, in many indus-
tries, firms are increasingly locked into a state of
network interdependency, whereby firms’ innova-
tion strategies must take account of the comple-
mentary innovation activities of others in the
network. Innovation, in such contexts, is often
R&D Management 36, 1, 2006. rBlackwell Publishing Ltd, 2006. Published by Blackwell Publishing Ltd, 67
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
driven by those firms who configure the network
in order to gain and maintain control over the
dispersed value-generating activities. However,
we suggest that such control is delicate and
industry dynamics can upset the roles and posi-
tion of such network configurers.
The paper moves on to investigate the nature of
network configuration through the study of the
evolution of three innovation networks in the
international fibre industry. By utilising a multi-
ple-case-study methodology, the key activities,
decisions and motivations of central firms operat-
ing in each network are explored. The discussion
shows how network configuration and control
was enacted. Critical explanations for the ob-
served development and configuration of the net-
works are uncovered, presented and discussed.
The paper concludes by drawing out implications
for managers operating in such contexts and
offers suggestions for further research in this area.
2. External collaboration for product
innovation
Scholars (Biemans, 1995; Tidd et al., 1997) have
argued that inter-firm collaboration provides a
means of managing some of the more complex
aspects of product innovation. While the short-
term collaborative benefits of risk and cost-
cutting and increased speed to market are
well-recognised (Lorange and Roos, 1991; Rice,
1991; Gugler, 1992; Bruce et al., 1995), the longer
term and more strategic contribution of colla-
boration to the innovation process has recently
been articulated. This includes greater intellectual
depth, opportunity scanning, competence en-
hancement, value-added solutions or worldwide
reach (Powell et al., 1999). It is argued that
collaboration is particularly useful in industries
such as ICT and biotechnology where there is
rapid obsolescence of technology and there are
significant barriers to accessing these technologi-
cal developments as a result of cost and/or own-
ership factors (Shan et al., 1994, Bruce et al.,
1995; Hinterhuber, 2002).
Other authors (Hamel et al., 1989; Perks, 2000)
have highlighted the benefits of sharing not only
technological developments but also other re-
sources, such as skills, knowledge and informa-
tion about competitors, customers, suppliers and
markets. Companies that are able to bring to-
gether complementary competencies are likely to
be in a better position to take advantage of
developments and achieve competitive advantage
(Shan et al., 1994). In particular, the value of the
dyadic relationship between manufacturer and
customer has received most attention in colla-
borative product innovation research. Integrating
customers in the innovation process can provide
considerable value through their input into the
generation of product ideas, information about
user requirements, comments on new product
concepts, assistance on development and testing
of prototypes, and assistance in diffusion (John-
sen and Ford, 2000).
Inter-organisational collaboration for innova-
tion has, in recent years, taken on a global
imperative. From the demand perspective, the
supply of products, particularly in business to
business markets, is required on a global basis
(Rigby and Jeffery, 2001). Often this requires
collaboration with competitors, distributors or
customers in regions where, either the company
lacks the infrastructure to service customers’
requirements to the necessary levels, or is pre-
vented from doing so by local legal restrictions.
Critically, for the context of innovation, the
innovative capacity itself has become increasingly
globally dispersed. A wide range of innovation
activities are carried out across extended global
networks of partnering firms. Such inter-organi-
sational linkages can provide firms with a source
of global competitive advantage.
Despite the recent focus on collaboration as a
means of improving the innovation process, there
has been limited questioning of these apparent
benefits and little analysis of the disadvantages of
collaboration. Bruce et al. (1995), for example,
have shown their surprise that the possible risks
and costs of collaboration are less prominently
articulated, despite growing evidence that colla-
borations do not always meet the expectations of
the collaborating partners. In their study of
collaboration within the information and com-
munication technology (ICT) sectors, Bruce et al.
(1995) noted that, for many firms, the collabora-
tive process made product development more
costly, complex and difficult to control and man-
age. The investigation also highlighted the risk of
strategic information leaking to collaborators.
These findings support other work (Prahalad
and Hamel, 1990; Khanna et al., 1998; Kale et
al., 2000), which draws attention to the danger of
giving away core skills, knowledge and capabil-
ities to partnering firms. It appears that colla-
boration for product innovation can lead to a
reduction in the control held by one organisation
over the development process itself. This may
even lead to loss of control over the final product
Helen Perks and Richard Jeffery
68 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
or service resulting from this process (Lorange
and Roos, 1991). Such risks present a dichotomy
for firms as they search for appropriate mechan-
isms to collaborate and share knowledge with
third parties, while simultaneously seeking to
maintain control over the innovation process.
3. Innovation networks
An emerging body of research recognises that
most product innovation relationships are not
created and developed in isolation, but are part
of a broader context of a network of interdepen-
dent relationships (Ha
˚kansson and Snehota,
1995; Biemans, 1999; IMP Group, 1999), from
which organisational members can access and
exchange varied resources (Bower, 1993). Within
this view, each relationship is connected to some
other relationship and cannot be understood if
these connections are disregarded. An important
characteristic of the network approach relates to
the definition of the boundary of a company.
Ha
˚kansson and Snehota (1995) note that while
a company is a clearly defined unit from a legal
point of view, from a resource and activity point
of view it is not: ‘The activities of the company are
part of activity chains and its resources are parts
of greater resource constellations . . . the company
is part of a context with activities and resources
which cannot be isolated from those of important
counterparts . . . a company should not be seen as
an island, but as part of a mainland network’
(Ha
˚kansson and Eriksson, 1993, p. 8). Cooke
(1996) suggests that a firm’s ability to develop
network relationships with others is based on its
existing relationships. Hence opportunities to
form linkages reflect prior patterns of inter-firm
relationships within a network (Ahuja, 2000).
Proponents of the network approach argue that
it opens up a new way of conceptualising compa-
nies within markets and offers new perspectives
on some traditional problems of product devel-
opment (Zajac and Olsen, 1993; Ha
˚kansson and
Snehota, 1995). Recent research has suggested
that a firm’s network position might explain its
performance outcomes more accurately than its
market position (Gulati et al., 2000). This lends
support to the adoption of a relational rather
than competitive-based approach to examining
firm behaviour. Network experience, articulated
in terms of knowledge of how to collaborate,
along with new knowledge gained through the
collaboration process itself, has also found to be
positively linked to the rate of innovation (Powell
et al., 1999).
Evidence has shown the importance of diversity
in types of actors within the network and its
impact on innovativeness. Kash and Rycroft
(2002) found that partner types are related to
the type of innovation occurring. This develops
previous work which indicated that incremental
innovators rely more heavily on customer invol-
vement (Biemens, 1991), whereas more advanced
innovators interact more frequently with other
collaborators, such as universities (Hausler et al.,
1994). Further, Conway (1995) showed that inter-
mediaries, such as trade/professional associations
and public bodies aimed at advancing innovation,
have a positive impact on developing innovation
networks. In particular, such bodies are effective
in facilitating informal relationships which are a
strong basis for the development of network
relationships (Cooke, 1996). Science partners (in-
cluding research and technical institutions) simi-
larly contribute to the development of informal–
personal networks and enable firms to access and
assimilate knowledge outside their normal busi-
ness systems (Liyanage, 1995; Verspagen, 1999).
4. Innovation inter-dependency
The incorporation of a contextual approach to
understanding networks and innovation furthers
the network perspective and begins to develop the
conceptualisation for this study. Industry sector
dynamics have been shown to influence the range
and nature of collaborative opportunities open to
firms seeking to innovate (Gulati et al., 2000). A
number of scholars have documented the value of
networking activity to innovation in a range of
industries where knowledge is distributed across
organisations and products are becoming increas-
ingly modular (Baldwin and Clark, 2000). In
empirical studies of the biotechnology and phar-
maceutical sectors (Gemser et al., 1996; Baum et
al., 2000), increased innovation rates were linked
to networked R&D. Gemunden et al. (1996), in
studying networking effects in six high tech in-
dustries, found that firms using networking were
likely to have 20% more product improvements
than firms that failed to network. An emerging
characteristic of many of these industry sectors is
the interdependency of firms, the interlocking
nature of the products they develop and the
shifting source of innovation in the industry.
Firms operate under high levels of innovation
inter-dependency across the network. They
become locked into a situation where their in-
novation strategies must take account of the
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 69
complementary product development activities of
others in the network. Firms are then tied to the
innovations of others. For example, in the ICT
industries final products are a bundle of compo-
nents, technologies and architectures. Innova-
tions from Intel and Microsoft (operating at an
upstream component level), for example, strongly
influence and shape the product development
directions of other firms in the IT industry
(Gawer and Cusumano, 2002).
5. Innovation network configuration
Questions then develop over how firms should
position themselves within such inter-dependent
networks and what kinds of network configura-
tions advance innovation. The literature is am-
biguous and inconclusive over what constitutes
successful or optimal network configurations for
innovation. Shan et al. (1994) propose that the
number of collaborative relationships within a
wide network is positively related to innovation.
Conversely, closed networks appear to act as a
more effective conduit for innovation than open
ones (Coleman, 1988). Others propose that all
firms should be tied to the focal actor (Brass and
Burkhurdt, 1992) who effectuates control over the
network. In the context of innovation inter-
dependency at an industry level, described above,
it is suggested that value created in the industry
can be configured by powerful focal actors, con-
ceptualised as orchestrators (Hinterhuber, 2002)
or platform leaders (Gawer and Cusumano,
2002). Such firms purposefully develop and co-
ordinate a wide range of network partners to
bring about innovation in the industry. In addi-
tion, by integrating diverse and dispersed activ-
ities throughout the network, new markets may
effectively be created. Further evidence points to
the superior innovative and financial performance
of those focal firms who creatively manage the
extended network of firms across the industry
(Hinterhuber, 2002). Such firms, who drive net-
work innovation, may effectively manipulate the
network and develop control over the value-
creating capacity in the industry.
Innovation network configuration, in this
study, is conceptualised as the shaping and man-
agement of the firm’s position in a network in
order to access and mobilise critical knowledge
for innovation which resides within the network.
The literature suggests that the most common
reason for collaborating in an innovation net-
work is to gain access to new or complementary
knowledge, primarily in the areas of technical
knowledge and commercial knowledge of markets
or customers (Cooke, 1996; Ahuja, 2000; Coles
et al., 2003). We propose that firms seek to
configure and manipulate their network to access
and retain such knowledge dispersed across the
industry network. By doing this, firms can more
effectively develop a role as enablers of innova-
tion across the network and control the direction
of the innovation.
Previous research suggests that a network can
create a strong identity and co-ordinating rules that
may become superior to the organisation itself in
its ability to create and disseminate know-
ledge (Dyer and Nobeoka, 2000). Processes to
form such an innovation network may be emergent
(developing from environmental changes and com-
mon network member interests; Conway, 1995).
However, in many cases, the network is triggered
or engineered by a key entity (Doz et al., 2000).
This firm may actively recruit network members
and develop, over time, a central position in leading
and configuring the network towards innovation.
The perspective of the focal actor, seeking to
exploit and configure the network to advance
innovation, is adopted in this study. However,
we also suggest that such a role can be tenuous.
Network configurations change and adapt ac-
cording to the requirements and capabilities of
the partners within it and shifting industry con-
texts (Kash and Rycroft, 2002; Koch, 2003).
Continuous innovation at all levels of the indus-
try network can alter the drivers of demand and
destabilise network positions. Network configura-
tions are dynamic and firms make choices and
adapt, over time, their management of such con-
figurations. Firms must orchestrate shifts and
manage and align their network position to
changes over time. Such dynamism is an impor-
tant feature of this study.
6. Research objectives and questions
This study seeks to further our knowledge of the
dynamics of global networks for innovation. The
empirical part of this research builds on and
further develops existing studies on the config-
uration of innovation networks (Gemunden et al.,
1996; Kash and Rycroft, 2002; Koch, 2003).
Specifically the research aims to explore the
different strategies that firms, in conditions of
innovation inter-dependency, use to configure the
network in order to access and mobilise critical
knowledge for innovation. Such strategies may
Helen Perks and Richard Jeffery
70 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
culminate in different configurations and different
innovation outcomes.
The research focuses on the actions and beha-
viours of the focal firm seeking to innovate within
a network. It unravels the dynamics of the parti-
cular mechanisms by which such key entities se-
lect and manage extended networks, over time,
throughout the innovation process. It further
aims to explore the drivers and explanations
behind observed firm behaviour in this context.
We suggest that the configuration of networks is
contingent on the industry context, the firm and
its network partners. We aim to explore and
unravel the nature of these contingencies.
The following specific questions will be ad-
dressed: How do innovation networks operate in
practice? How do organisations configure indus-
try networks in the innovation process? How do
they influence the direction of the innovation?
How and why does the shape and configuration
of the network change over time? What are the
underlying mechanisms which explain and drive
such approaches?
7. Research method
The study adopts a multiple case-study approach.
A case-study methodology enables the researcher
to maintain the complexities and contextual con-
tingencies in which the firms and the phenomena
under study are embedded (Yin, 2003). This is
particularly pertinent for network research where
it is difficult to isolate firms from the messy and
complex realities in which they operate (Das and
Teng, 2000). A case-study approach also allows
exploration of the evolutionary processes by
which the networks develop.
In determining the cases to be studied for this
research, the authors sought a common industry
setting which displays traits of innovation inter-
dependency among its players. The fibre industry
was chosen as an appropriate industry context. This
sector is undergoing structural change that is de-
manding novel practices from its players. The fibre
industry is characterised by a complex and multi-
level supply chain. It is experiencing high levels of
inter-dependency in innovation, along with shifting
nuclei of sources of innovation and power bases.
The researchers identified three innovations
developed throughout a network within the fibre
industry. The innovations displayed differences in
the way they evolved and had different outcomes,
permitting the exploration of divergent processes
within the chosen industry sector. The upstream
fibre company in each network was tracked as the
pivotal configurer and main unit of analysis.
These organisations were embedded in a complex
extended network leading, in each case, to the
development of a new fibre for the textile industry.
Key features of the central firms, the networks and
fibre innovations studied are given in Figure 1.
8. Data collection methods
The data collection for the study was completed
over a period of 3 years in the latter half of
the 1990s. As the research sought to explore the
perspectives of all companies involved in the
network for each innovation, the range of com-
panies engaged in each of the three fibre innova-
tions was identified. One of the authors had
extensive knowledge of and access to companies
in this sector. This helped the identification and
access to those firms, ranging from the originating
Central Firm Company Characteristics Nature of
Innovation
Network configuration
by focal firm
Company A International speciality
materials company,
operating in coatings,
performance materials,
packaging, fibres and films
Radical:
Development of
new generic fibre
type
Network configuration
outsourced through dyadic
relationship with third
party
Company B International chemical,
materials and energy
company operating in a
wide range of markets
including chemicals and
textiles
Incremental:
Line extension of
existing fibre
Centralised network
configuration by focal firm
Company C Chemical company
operating in fibres and
packaging markets
Incremental:
Development of
new biodegradable
fibre
Restricted network
configuration by focal firm
Figure 1. An overview of the three fibre innovation networks.
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 71
fibre company to downstream garment retailers,
involved in each network. Multiple in-depth and
semi-structured interviews were held with key
respondents, representing marketing, R&D and
senior management, in these companies. Inter-
views were carried out in Western Europe, North
America and the Far East. Innovation A was
explored as the key case and in most detail.
Supporting documentation and industry data
were also obtained. The number and spread of
interviews are shown in Figure 2.
9. Innovation shifts in the fibre Industry
Globally, 32.7 million tonnes of man-made fibre
was produced in 2000, rising from 1.7 million in
1950 (Johnson, 2001a). Apparel accounted for
half of all fibre consumption, with interior textiles
and carpets taking a 25% share. Other key
applications include medical/hygiene (such as
cotton wool and wipes), automotive/transport
(e.g. carpets, seat belts) and technical uses (e.g.
packaging). Worldwide, man-made fibre use is
expected to rise to 80 million tonnes by 2050.
The textile and clothing industry has a complex
structure. There are up to seven processing stages
between chemical fibre intermediate companies
and end-users in both industrial and consumer
markets. Yet, from the early commercialisation of
synthetic and regenerated fibres in the 1950s and
1960s to the mid-1970s, it was the fibre producers
who held much of the power across the whole
industry supply chain. At this time, companies
such as ICI and DuPont introduced and commer-
cialised, in high volumes, new fibres such as
rayon, nylon and polyester (Albrecht, 1999; John-
son, 2001a,b). Demand for fashion from consu-
mers was very low, with most markets demanding
cheap, basic garments in a limited range of
colours. In many cases the fibre companies had
strong links with the larger clothing retailers.
They were able to command pull-through de-
mand for their fibres and often capitalised on
their strong position in the chain to expand
downstream, through acquisition of textiles and
clothing manufacturers.
Towards the end of the 1970s and throughout
the 1980s the balance of power in the supply
network changed. Middle and lower market fash-
ion was created and fashion colours were adopted
within a system of seasonal styling and colour
changes. There were up to eight seasons in a year
and retailers emerged as the most important
decision makers, demanding a constant flow of
targeted garment innovation.
The economic recession in the late 1980s, coup-
led with stagnant consumer spending on clothing,
fuelled demand for classic garments and value for
money, rather than transient, and often expen-
sive, fashion garments. Retailers sought more
fundamental innovation, rather than simple col-
our changes and varied fabric weights. Quick
response and fast innovation was demanded.
The fabric manufacturers and converters took
on the most important decision-making roles in
innovation. They were able to exploit new fibres
and finishing techniques to produce true fabric
innovation. They also began to co-ordinate the
activities of spinners and finishers to respond to
the innovation needs of retailers and garment
makers. Collaborative new fabric development
programmes started to evolve.
The current early years of the 21st century are
proving to be a highly innovative period for the
global fibre industry. Textiles are becoming in-
creasingly specialised as more functions and prop-
erties are integrated into their structures. Factors
of renewable resources and recyclability are being
added to the important design considerations of
performance, protection and durability. Indeed,
ecological and environmental considerations are
now a major driving force in all fibre develop-
ments and initiatives. This is leading to dispersed
Number of interviews
Fibre
Innovation
Total Central
company
Yarn
pro-
cessors
Fabrics
weavers
and
knitters
Garment and
other
end-product
manufacturers.
Retailers
and
distributors
Other
stake-
holders
Innovation A 46 54 12 7
7
99
Innovation B 23 22 3 6 3
Innovation C 18 3- 6 5 3 1
Figure 2. Interview details.
Helen Perks and Richard Jeffery
72 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
and shifting power in the innovation supply chain
and much of the sources of innovation and power
reside with downstream garment retailers.
Figure 3 highlights some of the more important
innovations in the man-made fibre industry over
the past 50 years. After a significant number of
product innovations in the 1950s and 1960s there
were few further notable innovations until the
1990s. While this masks a myriad of incremental
innovations that took place during the interven-
ing years, it is apparent that the industry is
currently undergoing a highly innovative period.
10. Findings
Against this industry context, the following discu-
ssion is structured to show how three critical fibre
innovations were developed through the enact-
ment of a globally dispersed network. Focusing
on the upstream fibre company as the main unit
of analysis, the key activities, decisions and mo-
tivations of the critical actors in the network are
unravelled. The outcome, in terms of innovation
network configuration, is discussed. Explanations
for the outcome are drawn out and illustrated,
focusing on the underlying dimension of network
configuration. For each case study, a final table
identifies the network configuration outcome and
synthesises the key features of its development.
Greater depth and detail is provided for innova-
tion case study A which forms the focal case.
Innovations B and C develop the findings from
the first case study and provide further richness in
terms of contextual and behavioural differences.
10.1. Innovation case study A: innovation
through ‘outsourcing’ network
configuration to third party network
10.1.1. Company A background
Company A, at the time of the fibre development,
was an international speciality materials company
with headquarters in the UK and annual sales of
d2 billion. It employed 30,000 people across 39
countries. It had five business areas, including
fibres and films products which accounted for
35% of turnover. Seventy-eight percent of sales
derived from outside the UK.
For 20 years before the development of the new
fibre, Company A’s product development process
focused on incremental improvements to existing
product lines. Technically oriented teams forged
strong relationships with first tier customers (i.e.
yarn companies) and, less frequently, collabo-
rated with second tier (i.e. weavers, knitters)
process companies in the supply chain. The com-
pany used these relationships to identify on-going
opportunities to refine and improve existing pro-
ducts. The company had minimal influence over
down-stream processors (i.e. fabric finishers, gar-
ment manufacturers), retailers or end-users, who
were considered distant and far-removed, and
developed and promoted the fibres as they
wished. However, unlike many fibre companies,
Company A was able to develop some weaker ties
to external bodies, for example universities and
consultancies. This provided a low-cost access
route to downstream market knowledge.
10.1.2. The innovation: Cross
The impetus for the development of the new fibre
innovation (called Cross), considered in this case
study, came in the 1980s. By the late 1970s
Company A recognised that the product life
cycles of its existing products had reached matur-
ity. Moreover, environmentally friendly produc-
tion techniques were becoming increasingly
important to the European consumer and, in
turn, to the textile and clothing supply chain.
These factors led to the emergence of two key
strategic thrusts: firstly to find a means of redu-
cing the impact of the cellulosics industry on the
environment; and, secondly, to achieve a new and
radical direction in output to complement the
Year
2000 Cargill Dow launch NatureWorks, a PLA fibre made from corn
1997 Shell starts production of PTT polymers in USA
Lenzing launch “Lenzing Lyocell”
1992 Company A launches ‘Cross’
1976 Gore markets its “Gore-Tex” membrane
1973 DuPont launches Kevlar
1962 DuPont begins production of Nomex
1959 DuPont launches a stretch fibre “Lycra”
1954 Bayer starts producing Dralon
Hoechst introduce a polyester fibre “Trevira”
1952 Rhodia-ceta AG starts producing nylon in Frieburg
Innovation
Figure 3. Key fibre innovations 1950–2000.
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 73
company’s existing range of speciality cellulosic
fibres.
In 1978 Company A began an intense investi-
gation into the solvent dissolution of cellulose
from wood pulp. From the early results of this
project – codenamed ‘Origin’ – it was clear that a
radically new fibre technology was emerging. By
1988 a new basic fibre had been developed and
given the name Cross. Cross was a new generic
fibre type that includes fibres made from cellulose
by a solvation process as opposed to viscose and
modal, which are produced by a reactive process.
This has the advantage that during the produc-
tion of Cross, the solvent can be recycled for
continuous reuse with only very small losses. This
makes the process viable from an economical
perspective and means that there is little release
from a lyocell plant into the environment. Cross is
made from wood pulp and is therefore biodegrad-
able. Fabric produced from Cross has a soft
lustrous hand and high wet strength, unlike com-
peting fibre viscose, which has poor washability.
However, it took a further 14 years to take the
new fibre, and the technology needed to produce
it, from conception to full-scale production. The
sequence of events during the development of
Cross are shown in Figure 4.
The stages of the innovation development pro-
cess are described below. The discussion focuses,
in particular, on an analysis of external network
development in the 1990s.
10.2. Innovation network configuration
10.2.1. Idea generation stage: internal network
development
The early conception stage of the innovation
process was dominated by the use and develop-
ment of internal R&D networks. Whereas formal
linkages between technical departments were
strong, ties between R&D and marketing, how-
ever, were weak. Close internal R&D networking,
during the early phases, helped reduce uncertain-
ties and speed up the product conception stage.
Through this early internal networking approach,
the R&D team was able to quickly gather and
analyse comprehensive and relevant information
held by the company. This enabled them to dis-
card a number of possible approaches and iden-
tify previous unused research that could form a
basis on which to develop an R&D action plan.
10.2.2. Development stage: network configuration
by the Japanese Kai
External network linkages occurred at the devel-
opment stage of the innovation process. The firm
needed access to technical and, particularly, mar-
keting resources and capabilities, unavailable or
insufficient in-house, in order to move to the
development of a commercially viable fibre.
The Japanese textile industry was selected for
initial yarn and fabric development. The develop-
ment of Cross in Japan appeared attractive to
Company A. Japan was seen to hold a collection
of the most sophisticated processor companies,
best equipped to overcome the processing pro-
blems and develop the finishing and dyeing cap-
abilities of Cross. Japanese processors also had
considerable experience of enzymes (which are an
essential chemical in the processing of Cross).
This expertise resulted from the banning of for-
maldehyde in Japan during the early 1980s,
which, in turn, encouraged the use of enzymes
as an alternative process chemical. Company A
also felt that Japan’s extensive R&D capabilities
were also accompanied by expertise in controlling
the supply chain in its entirety.
Many processors in the Japanese textile supply
chain had also expressed, to Company A, a keen
desire to be involved in any new fibre develop-
ments. This high level of interest resulted from the
fact that most new fibre developments in Japan
were controlled by vertical operations, which
effectively blocked any collaboration with down-
stream processors who were not part of these
groups.
The drivers and objectives for the networked
Cross development from the perspective of Com-
Year Stage of Development
1970s Idea and conception stage: Screening process to find new cellulosic process
1980s First crude prototype developed: first pilot line capable of producing 100kg/week of fibre
1990s Pilot line scaled up to 1 tonne/week and establishment of initial network of development
partners in Japan
First cross pilot plant operational (40 tonne/week)
Fibre launched in Japanese apparel market
Development networks established in USA and Western Europe.
Plants commissioned in USA and Western Europe
Figure 4. Sequence of events in the development of ‘Cross’.
Helen Perks and Richard Jeffery
74 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
pany A and the Japanese Kai are highlighted in
Figure 5.
The product development strategy for Cross in
Japan relied on a new networking system under
the guidance of a Tokyo-based planning com-
pany. The Managing Director had previously
worked for a large trading company with large
textile interests, where he had gained extensive
experience in the operation of Japanese textile
supply chains. Two spinners, two weavers, one
knitter and four dyers and finishing companies
were selected, based on their technical and com-
mercial capabilities. The main aim of these com-
panies was to create methods to exploit and
develop the characteristics of Cross yarn and
fabric. This group of processors, together with a
small number of trading houses, became known
as the Cross Kai. The Kai had control over the
development of new yarns, fabrics and garments
and their subsequent pricing. The number of
companies involved in the Kai was limited to
ensure that quality was controlled and exclusively
maintained. An overview of the structure of the
Japanese Cross network, which emerged, is out-
lined in Figure 6.
All parties in the Kai were well-established
market focused companies with efficient internal
networks. Company A was able to take advan-
tage of the superior multi-level technical and
marketing capabilities of Kai without further re-
source commitment. Without these resources, the
early successful development of Cross by the Kai
would have been more problematic. The Kai’s
control and development of the network per-
mitted Company A to focus on core activities
without a need for capability enhancement. Com-
pany A was hence able to develop strong link-
ages with the Kai and gave the Kai freedom to
Company A Japanese Kai
Take advantage of an apparent opportunity Use the new fibre to solve existing problems
Development and commercialisation of a new
fibre
Extension of product line; take advantage of
leading edge developments
Shortening of overall development time Assure quick return from involvement with
network
Global spread of production Target home market and selected export markets
Move to mass market Retain high added value in home market niches
Figure 5. Drivers and objectives of the networked Cross development for Company A and the Japanese Kai.
Company A
(Fibre)
Spinner Co-ordinating
trading house
Knitter
Trading Houses
(four)
Apparel
Manufacturers
Consumers
Two
Weavers
Four
Dyers & finishers
Prototype
Fibre
Retailers
Prototype
Woven
Prototype
Yarn
Prototype
Yarn
Prototype
Yarn
Knitted
fabrics
Woven
Fabrics
Fabric
KEY
Formal information flows
Informal information flows
Prototype product flows
Prototype
Garments
Tokyo based
Planning
Company
Product
Planning
Figure 6. Map of Japanese ‘Cross’ Network (1998).
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 75
develop the network in its own manner. Company
A felt that critical innovation decisions were
hence taken by those firms with appropriate
experience and knowledge Indeed, the prior
knowledge and contacts of Kai allowed speedier
decision making and faster time to market for the
innovation. However, the Kai grew increasingly
powerful in the direction of the fibre network. For
example, the Kai proved reluctant to develop
blended products which may have opened up
new market segments for the fibre.
10.2.3. Launch stage: international network con-
figuration
A small group of fashion designers, operating at
the exclusive end of the market, were integrated
into the Kai network, as the product moved closer
to a stage where it could be launched. This
strategy was adopted to create an exclusive image.
It was hoped that, over time, the product would
trickle down to lower priced, but higher volume,
product/market segments.
Indeed Cross was initially launched as a pre-
mium, luxury fibre in the Japanese market. It was
introduced exclusively into the upper consumer
brackets. This exclusive small volume positioning
aligned well with Company A’s limited produc-
tion capacity and the unique aesthetic qualities
(such as excellent drape and a soft handle) of the
fabric.
The first products of the Japanese Kai were
shown in a fabric exhibition in Tokyo in Decem-
ber 1990 and commercialised in 1991. Following
the commencement of full-scale production in the
USA, Cross was introduced into the American
market with a high profile fashion show. Cross
was launched in Europe 18 months later.
10.2.4. Innovation case study A: network outcomes
Garments developed from the Cross fibre showed
a high early market take-up. The success of the
Kai-developed garments arose out of exclusive
high-end market positioning, which aligned well
with limited fibre production capacity, and the
revolutionary nature of the fabric. In terms of
network configuration and position, Company A,
however, became increasingly distanced from the
activities of the Kai and some of the linkage pins,
most notably the Tokyo-based planning com-
pany, became more central to the development
of Cross. A limited amount of the knowledge
being accumulated by the processors in the Kai
was fed back to Company A (Figure 7).
In essence, the company became removed from
key information and knowledge flows.
Ties with garment makers (downstream actors)
were also weak and some early problems occurred
at the garment manufacturing stage. Insufficient
care was taken in the handling of the fabric and
some garment makers attempted to use Cross in
garments for which the fabric was totally unsui-
table. While garment makers were part of the
network they tended to be ‘isolates’ and had little
input into the development process.
Further, Company A struggled to develop
linkages with critical wood pulp suppliers and
suffered from raw material costs which rose
dramatically. In the 12 months up to May 1995
spot prices for wood pulp rose from US$400 to
US$1,200 per tonne; a rise of 200% (Grimond,
1995).
Weak ties with downstream actors inhibited
Company A’s ability to understand and antici-
pate both market changes and the potential
extended benefits of its technology. For example,
Cross was incorporated into denim and garment
washed ‘peachskin’ products, which took advan-
tage of Cross’s unique attributes. However, these
markets were highly susceptible to fashion shifts.
In the second half of 1997 Cross sales fell sharply
as a result of a weak market for denim, falling
Asian demand and de-stocking by customers.
During 1997 the percentage of Cross sales used
in denim products fell from 80% to 20%.
10.3. Innovation case study B: innovation
network centrality through sub-
networks
10.3.1. Company B background
Company B is one of the most financially success-
ful chemical fibre companies operating within the
textile and clothing supply chain. It had global
Network configuration outcome:
Lack of control by focal firm over direction of innovation
Network development and control handed
over to Japanese Kai
Weak access by focal firm to critical
knowledge holders
Figure 7. Network configuration outcome and features of development of Cross (Company A).
Helen Perks and Richard Jeffery
76 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
sales of $30 billion in 2003, with R&D expendi-
ture of d1.3 billion in the same year. It has a long
history of fibre innovation and has commercia-
lised many of the most recognisable fibres avail-
able today.
Company B operates in the most complex
development network in the textile and clothing
industry. Multiple innovations are pursued for a
wide variety of end-uses, from lingerie and hosi-
ery through to tire-cord and airbags. In order to
remain a central player in the network, Company
B has developed an approach whereby it selects
from across the network, at different levels, and
builds ‘sub-networks’ within its wider set of
relationships. It chooses a small group of core
companies and dedicates a focused cross-func-
tional in-house team to the specific development.
10.3.2. The innovation: Perfo Strata
In 1989 Company B acquired the nylon fibre
interests of a UK chemical company. One of the
main types of nylon produced by the UK com-
pany, at that time, was called ‘Perfo’. Company B
took this brand and initiated a period of innova-
tion through a process of product extension, with
new variations to meet new end-uses. In 1997
Company B introduced ‘Perfo’ Strata. This is a
bi-spun product which creates a two-tone colour
effect combined with a soft handle. This was
originally developed for the circular knitted fabric
industry and was then extended into knitwear and
woven fabrics for products such as casual and
business trousers.
10.3.3. Innovation network configuration
In the development of ‘Perfo’ Strata, Company B
selected a leading dye stuff manufacturer, an
international retailer, two circular knitted fabric
producers and a knitwear manufacturer. From
the outset of each of these relationships, clear
intellectual property guidelines and expected two-
way information flows were specified by Com-
pany B. In addition, its long-term objectives were
shared with the development partners at an early
stage of the product development.
The development partners were compelled to
grant Company B exclusive rights to any intellec-
tual property discoveries. The need to maintain its
identity and the ‘Perfo’ brand name was a core
element of the network configuration. However,
despite the fact that Company B has many
patents on the downstream processing of ‘Perfo’
Strata, it was largely willing to allow its customers
to use these and other patented technologies
freely as long as they used Company B’s products.
The IP was agreed on in such a way that it
encouraged and promoted the use of ‘Perfo’
Strata in many applications. Benefits for the
downstream processor were further achieved
through the extensive technical and promotional
support provided by Company B. Such support
activities were provided not only through direct
technical and marketing support but also market-
ing initiatives which helped the network as a
whole. This network organisation ensured perfo’s
route to the end-users was clearly planned.
Indeed, Company B is one of the few fibre
companies to pursue a strategy of consumer
branding and promotion. However, support in-
itiatives were closely linked to well-defined and
mutually understood performance levels for each
of the network partners. While this helped achieve
demand-pull from the market it also ensured
commitment to Company B and the sub-network.
The risk of losing access to such support and
promotion discouraged processors from using the
knowledge and technologies, gained from the
processing of Company B fibres, for the proces-
sing of competitors’ fibres. Heavy usage of down-
stream promotion and consumer branding also
demonstrated reciprocity in investments to part-
ners and encouraged commitment and stability.
Network stability was also achieved by effective
selection of the right development partners. There
was emphasis on working with companies which
shared its goals and culture. By working with
companies which had experience of operating in a
network controlled by Company B, all parties
built up, not only confidence in each other, but
also ‘network experience’. Those companies
which worked well within these networks became
favoured partners and were more likely to be
chosen to work in future networks. These rela-
tionships allowed Company B to facilitate and
control adequate flows of information with the
confidence that all strategic information would be
fed back to it.
10.3.4. Innovation case study B: network outcomes
By 2003 there were indications that sales targets
were being met and plans to move the fibre into
other applications were under development. Cri-
tically, Company B retained centrality in the
network. It remained a central configurer of the
network, even though much development took
place further down the supply chain. Further-
more, exploitation of its central network position
allowed Company B to adapt to market and
industry changes and respond to emerging mar-
kets. Access to downstream knowledge enhanced
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 77
its ability to open up new markets for the fibre
(such as knitwear) well ahead of other fibre
companies. Indeed, knowledge about potential
markets, emanating from partner companies in
its wider network, facilitated entry into the bath-
room floorcovering market with the Perfo fibre
(Figure 8).
10.4. Innovation case study C: innovation
network control through a focused
network
10.4.1. Company C background
Company C was formed in 1997 as a joint venture
between one of the main global producers of
chemicals and a major international grain produ-
cer. The company was initiated after the parent
chemical company had developed the ability to
commercially ferment corn starch to produce a
polylactic acid (PLA). The innovation was given
the name ‘EarthFibre’. The company is an inde-
pendent company based in the USA. By 2002,
over US$750 million had been invested in re-
search, development and manufacturing capabil-
ities to bring EarthFibre to market.
10.4.2. The innovation: EarthFibre
EarthFibre is a natural alternative to oil-based
synthetic polymers and is exploited in end-uses
such as packaging, films, bottles and textile fibres
(both filament and staple). This is a major in-
novation in the textile and clothing industry.
Company C has been able to benefit from the
industry-wide learning outcomes of the develop-
ment of other fibres, by other firms, such as Cross.
10.4.3. Innovation network configuration
Company C focused on relationships with a small
number of downstream processors in each specific
market segment and carefully researched the
capabilities of its partners. During the early stages
of development, this focused strategy allowed the
best use of manpower and available polymer.
Two early relationships defined the company’s
approach. In 2000, Company C entered into a
relationship with a Japanese processor. However,
no up-front intellectual property agreements were
reached and the Japanese company took out
patents on almost all subsequent product devel-
opments incorporating EarthFibre. This severely
restricted any technology transfer from Company
C to other collaborators within the Asian region.
Following on from these early experiences
Company C subsequently placed high emphasis
on the control of the network and the information
flows which resulted from this. It set up a new
intellectual property system during the establish-
ment of a relationship with a Korean processor.
The Korean company retained all intellectual
property from its own development work on
EarthFibre, but then licensed IP out to other
processors of EarthFibre on demand by Com-
pany C. A licence fee was subsequently collected
by Company C in the form of a surcharge on the
chemicals delivered to the Korean company. In
addition, in order to avoid leakages or misuse of
information, Company C established a number of
parameters applied to all new development rela-
tionships:
Heavy investment, within each product/mar-
ket segment, into researching the technical
and commercial capabilities of all potential
development partners, in order to ensure that
each has the prerequisite capabilities, as well
as a long-term desire to work with a renewable
raw material.
Early establishment, once companies were
selected and initial discussions taken place,
of relatively stringent measures to ensure that
all learning outcomes from any technical work
are shared.
Adoption of a market development strategy
for the development of ‘EarthFibre’ whereby
each potential market is prioritised and
tackled in turn, to ensure resources are not
spread too thinly.
Company C had regular updating meetings
with all development partners where all informa-
tion was shared. Furthermore, a common mes-
sage was presented to the wider textile industry.
This was based on a global advertising campaign
in both the trade press as well as national pub-
Network configuration outcome:
Focal firm control over direction of innovation
Development of
network experience
of members
Investment in
support activities
Development of
knowledge-sharing
mechanisms
Figure 8. Network configuration outcome and features of development of PerfoStrata (Company B).
Helen Perks and Richard Jeffery
78 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
lications. However, early on, some leading com-
panies were unwilling to work with Company C if
all learning outcomes had to be shared. They also
felt constrained by the agreements in place and
lost confidence in their ability to experiment with
the new fibre.
10.4.4. Innovation case study C: network outcomes
Company C developed a number of small net-
works focused around specific product markets. It
invested heavily in identifying specific markets
and potential people to work with in each of
these segments. This selection was not only based
on the technical abilities and access to markets of
each partner but, more importantly, their buy-in
to the brand concept which was being developed
around the fibre.
These ‘EarthFibre’ networks launched the fibre
with retailers in a number of global markets.
Indications from retailers suggest that this fibre
is likely to consolidate and grow its share of its
target markets. In addition, in 2003 Company C
announced further agreements with 85 brand
manufacturers, textile processors and other par-
ties to develop and promote ‘Earthfibre’. How-
ever, by concentrating on maintaining small
networks of companies, some members outside
these networks felt that Company C was missing
opportunities for cross fertilisation and develop-
ment of ideas (Figure 9).
11. Discussion
The three innovations in our study were effectu-
ated through a network approach. We suggest
that a network approach in industry sectors, such
as the fibre industry, is being driven by two
critical industry dynamics, which reflect broad
trends in innovation in global industries. Firstly,
the location of innovation value in the industries
is becoming increasingly fluid. Value, in terms of
critical knowledge for innovation, is either shift-
ing to new levels in the network or being created
by new industry entrants. Secondly, where there
are high levels of interdependency between indus-
try players, demand for one firm’s core product
can depend on an array of innovation at other
levels in the industry. This means that firms are
limited in what they can do on their own. Our
findings confirm that the limited competency base
of firms, coupled with the dynamics of innovation
value location across the industry, is driving firms
to develop relationships beyond their own bound-
aries to access critical value in the industry.
The innovations studied in this research ex-
plored three different network configuration
paths adopted by firms. The findings demonstrate
that the manner in which firms manage and
configure innovation networks can influence the
direction of the innovation and eventual innova-
tive capabilities of the original innovating firm.
Configuring and managing global innovation net-
works is a challenging task. Our findings have
shown how firms struggle to develop and manage
an array of mechanisms to do this.
From our findings we are able to develop a
typology of network configuration for innova-
tion, identifying three configuration characterisa-
tions. Figure 10 outlines this typology. It
identifies three configuration types, highlighting
their core features, and summarises the implica-
tions of each type for networked innovation
management, as drawn from our case findings.
These features and implications are subsequently
discussed in more detail.
11.1. Outsourced network configuration
Our findings suggest that the lack of network
management experience and low knowledge of
such value, residing at multiple levels in the
broader network, can encourage firms to establish
dyadic relations with known immediate or ‘next
tier’ companies only. Consideration of the ex-
tended network as a whole can be difficult for
firms embedded in their own base of knowledge
and patterns of relationships. There are advan-
tages to this approach. A next tier third party may
be best equipped and experienced to manage and
develop the original technology through to mar-
ket by leveraging its own networking capabilities.
This can occur through nurturing new, or exploit-
ing existing, networks with the appropriate criti-
cal linkages. Such was the case with Company A
Network configuration outcome:
Focal firm control over restricted direction of innovation
Development of
focused network
Restricted and specified innovation
information flows
Figure 9. Network configuration outcome and features of development of Earthfibre (Company C).
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 79
and the Kai in our study. However, a danger lies
in that the original innovating firm can find itself
with non-existent or weak relationships with the
more distant, yet pivotal, holders of innovation
value in the network. Such an approach can allow
the more experienced networked player to de-
velop a leading role in the innovation develop-
ment. Ultimately the originating firm may lose
control over the destiny and direction of the
innovation.
11.2. Centralised network configuration
A centralised network configuration suggests an
approach that emphasises investment in develop-
ment activities which support the network as a
whole. It appears that network centrality encom-
passes adopting a broader vision of innovation. It
involves recognising where the innovation value
resides in the network and developing capabilities
and mechanisms to understand and access such
value. However, critically, firms make efforts not
only to access the value, but to invest in the value
development activities across the network as a
whole, for example through support or promo-
tional activities, as shown in innovation B in our
study. Such investment activities may not accrue
immediate dyadic benefits but appear to ensure
that knowledge flows are reciprocal and the net-
work develops successfully in the longer term.
The development of mechanisms by central enti-
ties to induce sharing of valuable knowledge and
rents between members appears to induce greater
performance of the network as a whole.
These findings further develop evidence from
existing studies (e.g. Dyer and Nobeoka, 2000;
Kogut, 2000), which confirm that developing
ways to share knowledge between network mem-
bers can more effectively advance the goals of the
network than dyadic relationships between mem-
bers. Our findings also suggest that maintaining a
position of network centrality may equip firms
with the ability to predict technological and
market changes sooner than their competitors.
This can allow such firms to develop a strong
influence over the direction of innovation in their
industries.
The critical issue here appears to be one of
capability extension. In our study, Company B
was able to gain control over the direction of the
innovation by extending, to some degree, its
capabilities into downstream and broader do-
mains, outside its original core base. This builds
on Gawer and Cusumano’s research in the ICT
sector (2002). By knowing more than you actually
make, firms are in a better position to understand
and assess the competencies of others in the
network and configure relationships in the net-
work more effectively.
11.3. Specified network configuration
Firms learn from others’ experiences in the in-
dustry. In our study, we have found that the fear
of losing control over the network can induce
firms to develop and manage networks through
tight operating specifications from the outset. Our
findings from Company C, which followed this
Network configuration type Core features Implications for networked
innovation management
Outsourced network
configuration
Dyadic relationship with
networked organisation
Weak ties to distant holders of
innovation knowledge
Risk of loss of control over
direction of innovation
Innovation direction driven by
collaborator
Centralised network
configuration
Development of central control
over extended network
Investment in indirect
capabilities to locate and access
distant innovation knowledge
Recipricol flows of innovation
knowledge
Ability to anticipate future
innovation shifts
Specified network
configuration
Specification of restricted
network membership and
knowledge flows
Entrenchment in network
structure
Ignorance of radical innovation
opportunities
Unilateral flow of innovation
knowledge
Figure 10. A typology of network configuration for innovation.
Helen Perks and Richard Jeffery
80 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
approach, suggest, however, that tightly dictating
network partners and their conditions, and spe-
cifying knowledge flows among them, may
achieve control but not centrality in the network.
By restricting and focusing the network, there is a
risk that new ideas emerging from potentially new
members of the network are ignored.
Our findings from Company C also indicate
that excessive control can lead to problematic
knowledge flows for subsequent innovation, par-
ticularly in the early idea generation phases. The
danger here is that as the core firm more deeply
specifies the processes of its network partners, it
not only stretches its managerial and technical
resources, but may turn the network into an old
style vertically integrated firm. This can lead to
entrenchment in the network structure and ignor-
ance of possible radical changes in the environ-
ment and products. Specified processes and
knowledge flows may allow only incremental-
style innovation and new product developments.
The data also suggest that failure to demonstrate
reciprocity in relationships can lead to negative
behavioural consequences in the network rela-
tionships, for example mistrust. Again this can
hamper effective knowledge flows for innovation,
as firms lack the motivation to pass on new ideas
and knowledge.
12. Conclusions and managerial
implications
Understanding the mechanisms and required cap-
abilities to manage innovation through a global
network is becoming a topical issue. This impera-
tive is particularly acute for those managers
operating in international sectors where the va-
lue-generating activities are shifting and becom-
ing increasingly inter-dependent. Managers face
tough decisions linked to the level of resources
they are willing and able to invest in such net-
works and how far they want and need the firm to
stray from its core capabilities.
Our study has gone some way to unravelling
the way firms manage and configure innovation
networks and some key implications for manage-
ment action are drawn. Firstly, our findings
suggest that, in order to access meaningful knowl-
edge from other domains in the network, firms
must enlarge the scope of their capabilities. In
other words, firms may need to develop a degree
of internal knowledge of these areas in order to
leverage value from them. This may require en-
hancing knowledge in areas downstream or up-
stream of the core base. In particular, for many
industry sectors, it appears critical to stimulate
demand through involvement in downstream ac-
tivities where the critical value – end use applica-
tions – are developed. This can help prime and
direct the market towards technologies and ap-
plications where the core firm operates. The
extension of capabilities requires careful deci-
sion-making about where investments in new
knowledge bases should be made.
Secondly, our finding suggest that managers need
to find appropriate levels of investment and me-
chanisms to help evolve the innovativeness of net-
work members. Too little involvement can mean
losing control, whereas too much can mean restrict-
ing the innovative capabilities of partners. Too
much control over what other firms do is likely to
hamper their innovative behaviour. In order to do
this, managers may be impelled to provide re-
sources (be it technical, commercial/marketing or
financial) to support the development of innovative
capabilities. For managers, this may involve, for
example, developing an ‘innovation-enabling’ role,
whose remit is to advise on the direction and level of
investment in the network members’ activities.
Managers may further need to provide suffi-
cient incentives for network partners to want to
work with and build upon the technology or
innovation emanating from the core firm. Other
firms in the network may not naturally have
incentives to follow a firm’s lead – it has to be
worth their while. This relates to decisions about
the sharing of knowledge and information about
the firm’s own technologies and products. Firms
may need to disclose enough information about
the core technology to entice external firms to
develop it. The management of intellectual prop-
erty sharing is key. Exposing too much informa-
tion may engender imitation and loss of control
over valuable sources of competitive advantage.
But disclosing information can help foster exter-
nal innovation and appears to act as a key
mechanism in safeguarding the third party’s free-
dom to innovate.
Finally, our study findings suggest that man-
agers need to look at the whole innovation net-
work and see how the pieces fit together. They
need to pay attention to the business incentives
for all involved. Being concerned with the com-
mon good of the network and expanding value
for everyone is a difficult perspective to adopt.
Traditional business incentives reward managers
on achievement of corporate goals yet a shift
towards focus on network performance metrics
is indicated.
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 81
13. Future research directions
While this paper has contributed to the under-
standing of how firms develop and manage a
network approach to innovation, it is clear that
our knowledge of global network configuration is
still emergent. While developments in conceptua-
lising networks are further advanced, more em-
pirical research is needed to strengthen its
foundation. In particular, more in-depth case
histories would enrich the field by mining the
details of the network management mechanisms
firms employ. Cross-industry comparisons could
unravel the industry contingent factors influen-
cing the development of innovation networks.
Finally, further studies might usefully extend
this research by empirically testing the scope
and contingent relationships between the key
network management mechanisms unravelled in
this paper.
References
Afuah, A (2000) How much do your competitors’
capabilities matter in the face of technological
change? Strategic Management Journal,21,3,
387–404.
Ahuja, G. (2000) The duality of collaboration: induce-
ments and opportunities in the formation of inter-
firm linkages. Strategic Management Journal,21,
317–343.
Abrecht, W. (1999) An amazing 50 years of fiber
development. Chemical Fibres International,50, 92.
Baldwin, C.Y. and Clark, K.B. (2000) Design Rules,
Vol. 1, The Power of Modularity. Cambridge, MA:
MIT Press.
Baum, J., Calabrese, T. and Silverman, B. (2000) Don’t
go it alone: alliance network composition and start-
ups performance in Canadian biotechnology. Strate-
gic Management Journal,21, 267–294.
Biemens, W.G. (1991) User and third party involve-
ment in developing medical equipment innovations.
Technovation,11, 163–182.
Biemans, W.G. (1995) Internal and external networks
in product development. In Bruce, M. and Biemans,
W.G. (eds), Product Development: Meeting the Chal-
lenge of the Design–Marketing Interface. New York,
NY: John Wiley and Sons.
Bower, D.J. (1993) New product development in the
pharmaceutical industry: pooling network resources.
Journal of Product Innovation Management,10,5,
367–375.
Brass, D. and Burkhardt, M. (1992) Centrality and
power in organisations. In Nohria, N. and Eccles, R.
(eds), Networks and Organisations. Boston, MA:
Harvard University Press. pp. 191.
Bruce, M., Leverick, F. and Littler, D. (1995) A
management framework for collaborative product
development. In Bruce, M. and Biemans, W.G.
(eds), Product Development: Meeting the Challenge
of the Design–Marketing Interface. New York, NY:
John Wiley and Sons.
Coleman, J. (1988) Social capital in the creation of
human capital. American Journal of Sociology,94,
95–120.
Coles, A., Harris, L. and Dickson, L. (2003) Testing
goodwill: conflict and cooperation in new product
development networks. International Journal of
Technology Management,25, 51–64.
Cooke, P. (1996) The new wave of regional innovation
networks: analysis, characteristics and strategy.
Small Business Economics,8, 159–171.
Conway, S. (1995) Informal boundary-spanning com-
munication in the innovation process: an empirical
study. Technology Analysis and Strategic Manage-
ment,7, 327–342.
Das, T.K. and Teng, B.S. (2000) A resource-based
theory of strategic alliances. Journal of Management,
26, 3, 31–61.
Doz, Y.L., Olk, P.M. and Smith Ring, P. (2000)
Formation processes of R&D consoria: which path
to take? Where does it lead? Strategic Management
Journal,21, 3, 239–266.
Dyer, J.H. and Nobeoka, K. (2000) Creating and
managing a high performance knowledge-sharing
network: the Toyota case. Strategic Management
Journal,21, 3, 345–367.
Freeman, C. (1982) The Economics of Industrial Innova-
tion. Cambridge, MA: MIT Press.
Gawer, A. and Cusomano, M.A. (2002) Platform Lead-
ership. Boston, MA: Harvard Business School Press.
Gemser, G., Leenders, M. and Wijnberg, N. (1996) The
dynamics of inter-firm networks in the course of the
industrial life cycle: the role of appropriability. Tech-
nology Analysis and Strategic Management,8, 439–453.
Gemunden, H.G., Ritter, T. and Heydebreck, P. (1996)
Network configuration and innovation success: an
empirical analysis in German high-tech industries.
International Journal of Research in Marketing,13,
449–462.
Grimond, M. (1995) Chemical costs corrode margins at
(Company A). The Times, 26 May. p. 21.
Gugler, P. (1992) Building transnational alliances to
create competitive advantage. Long Range Planning,
25, 1, 90–99.
Gulati, R., Nohria, N. and Zaheer, A. (2000) Strategic
networks. Strategic Management Journal,21,3,
203–215.
Ha
˚kansson, H. and Eriksson, A.K. (1993) Getting
innovations out of supplier networks. Journal of
Business to Business Marketing,1, 3, 3–34.
Ha
˚kansson, H. and Snehota, I. (1995) Developing
Relationships in Business Networks. London: Thom-
son Business Press.
Helen Perks and Richard Jeffery
82 R&D Management 36, 1, 2006 rBlackwell Publishing Ltd. 2006
Hamel, G., Doz, Y. and Prahalad, C.K. (1989) Colla-
borate with your competitors – and win. In Wit, B.
and Meyer, R. (eds), Strategy Process, Content,
Context. London: Thomson Business Press.
Hausler, J., Holn, H. and Lutz, S. (1994) Contingencies
of innovative networks: a case study of successful
R&D collaboration. Research Policy,23, 47–66.
Hinterhuber, A. (2002) Value chain orchestration in
action and the case of the global agrochemical
industry. Long Range Planning,35, 6, 615–635.
IMP Group (1999) An interaction approach. In Wil-
son, D. (ed.), Organizational Marketing. London:
Thompson Business Press. pp. 214–236.
Johnsen, T. and Ford, D. (2000) Managing collabora-
tive innovation in complex networks: findings from
exploratory interviews. In Proceedings of the 16th
Annual IMP Conference (CD-Rom), Bath.
Johnson, T. (2001a) Evolution of the european man-
made fibre industry. Chemical Fibers International,
51,4.
Johnson, T. (2001b) Repositioning in the man-made
fiber industry: Europe and America. Chemical Fibers
International,51, 10–11.
Kale, P., Singh, H. and Perlmutter, H. (2000) Learning
and protection of proprietary assets in strategic
alliances: building relational capital. Strategic Man-
agement Journal,21, 217–237.
Kash, D.E. and Rycroft, R. (2002) Emerging patterns
of complex technological innovation. Technological
Forecasting and Social Change,69, 581–606.
Khanna, T., Gulati, R. and Nohria, N. (1998) The
dynamics of learning alliances, competition, coop-
eration and relative scope. Strategic Management
Journal,19, 193–210.
Koch, C. (2003) Innovation networking between
stability and political dynamics. Technovation,24,
729–739.
Kogut, B. (2000) The network as knowledge: generative
rules and the emergence of structure. Strategic Man-
agement Journal,21, 3, 405–425.
Liyanage, S. (1995) Breeding innovation clusters
through collaborative research networks. Technova-
tion,15, 553–567.
Lorange, P. and Roos, J. (1991) Why some strategic
alliances succeed and others fail. Journal of Business
Strategy,12, 1, 125–30.
Perks, H. (2000) Marketing information exchange
mechanisms in collaborative new product develop-
ment: the influence of resource balance and competi-
tiveness. In Industrial Marketing Management,29,2,
179–189.
Powell, W., Koput, K., Smith-Doerr, L. and Owen-
Smith, J. (1999) Network position and firm perfor-
mance – organisational returns to collaboration in
the biotechnology industry. Research in the Sociology
of Organizations,16, 129–159.
Prahalad, C.K. and Hamel, G. (1990) The core compe-
tence of the corporation. In Wit, B. and Meyer, R.
(eds), Strategy Process, Content, Context. London:
Thomson Business Press.
Rice, V. (1991) Why teaming up is so hard to do.
Electronic Business,8, 30–34.
Rigby, D. and Jeffery, R. (2001) European markets for
technical textiles to 2010. In Proceedings of TechTex-
til Symposium 2001, Messe Frankfurt, Germany.
Shan, W., Walker, G. and Knought, B. (1994) Inter-
firm cooperation and start-up innovation in the
biotechnology industry. Strategic Management Jour-
nal,15, 387–394.
Tidd, J., Bessant, J. and Pavitt, K. (1997) Managing
Innovation: Integrating Technological, Market and
Organizational Change. New York, NY: Wiley.
Verspagen, B. (1999) Large firms and knowledge flows
in the Dutch R&D system: a case study of Philips
Electronics. Technology analysis and Strategic Man-
agement,11, 211–233.
Yin, R.K. (2003) Case Study Research: Design and
Methods (3rd edn). Beverley Hills, CA: Sage.
Zajac, E.J. and Olsen, C.P. (1993) From transaction
cost to transaction value analysis: implications for
the study of interorganizational strategies. Journal of
Management Studies,3, 131–145.
Global network configuration for innovation
rBlackwell Publishing Ltd. 2006 R&D Management 36, 1, 2006 83
