Strategic repositioning by means of alliance …...Strategic repositioning by means of alliance networks: The case of IBM Koen Dittrich (corresponding author) RSM Erasmus University,

This paper has been accepted for publication in Research Policy. Please cite this article in press as: Dittrich, K., Duysters, G. and De Man, A.P. Strategic repositioning by means of alliance networks: The case of IBM. Research Policy (2007), doi: 10.1016/j.respol.2007.07.002.

Strategic repositioning by means of alliance networks:

The case of IBM

Koen Dittrich (corresponding author)

RSM Erasmus University, Dept. Management of Technology and Innovation

PO Box 1738, 3000 DR Rotterdam, The Netherlands

Phone: +31 10 4082597

Fax: +31 10 4089014

Email: [email protected]

Geert Duysters

UNU-MERIT

Keizer Karelplein 19, 6211 TC Maastricht, The Netherlands

and

Eindhoven Centre for Innovation Studies (ECIS)

PO Box 513, 5600 MB Eindhoven, The Netherlands


Ard-Pieter de Man

Eindhoven University of Technology, Faculty of Technology Management

Eindhoven Centre for Innovation Studies (ECIS)

PO Box 513, 5600 MB Eindhoven, The Netherlands


K. Dittrich et al. / Research Policy xxx (2007) xxx-xxx


1

Abstract

This paper aims to show that alliance networks can play an important role in facilitating

large-scale strategic change projects. It focuses on the particular case of IBM, whose

radical re-direction from an exploitation strategy towards an exploration strategy was

realized by major changes in its network strategy. We show that by involving new

partners in the network and by loosening the ties with its existing partners, IBM

managed to transform from a hardware manufacturing company to a global service

provider and software company. The findings suggest that the traditional view of large

firms as being slow to adapt may not be valid because alliance networks can be used to

overcome inertia.

Keywords: Alliance networks; strategic change; exploration/exploitation; IBM.

1. Introduction

The strong upheaval in the number of newly established strategic technology alliances

has drawn the attention of academics and practitioners alike. There is growing

consensus in the academic and managerial literature that alliance networks do matter for

the innovative performance of companies (see e.g. Ahuja 2000; Hagedoorn, 1993;

Nooteboom, 1999; Owen-Smith and Powell, 2004; Powell et al., 1996; Rowley et al.

2000). However, the same publications seem not to clarify how such alliance networks

can act as catalysts of large-scale strategic change projects (De Man and Duysters,

2005). Strategic change is generally seen as something internal to the firm and not

directly related to the alliance portfolio of the firm.



2

This paper departs from the existing literature in the sense that it proposes that

external relations in a company’s alliance network can be instrumental in strategic

change processes as well. The main contribution of the paper is that we describe the role

of alliance networks in strategic change processes and support this view by studying one

of the most impressive strategic repositioning programmes in the history of business,

that of IBM. In this paper, we analyze how deliberate changes in the network structure

enabled IBM to reposition itself in an unprecedented manner. Although strategic

alliances have been studied extensively in the existing literature there is to the best of

our knowledge no single study devoted to the role of strategic alliance networks in

major company repositioning processes. Although some recent articles have focused on

issues related to alliances as means for achieving exploration or exploitation strategies

(e.g. Koza and Lewin 1998; Lavie and Rosenkopf, 2006; Rothaermel, 2001; Rothaermel

and Deeds, 2004), none of these has looked empirically and from a longitudinal point of

view at the overall impact of a firm’s network on its ability to transform its business

activities and competencies drastically.

Our view therefore extends strategic management theory in terms of including

alliance networks as a major repositioning tool. It departs from the classical view in

which large firms reposition themselves through internal restructuring processes or by

means of Merger and Acquisition activities and can also be seen as a solution to the

structural inertia problem which is widely discussed in organizational ecology theory

(e.g. Geroski, 2001; Guillén, 2004; Hannan and Freeman, 1984; Narula, 2002) and to

the slow adaptive nature of organizations as discussed in core competency approaches

(e.g. Bouchikhi and Kimberley, 2003; Johnson-Cramer et al., 2003; Markides and

Williamson, 1994) and evolutionary economic theory (Ballot and Taymaz, 1997; Louçã



3

and Mendonça, 2002; Suarez and Olivia, 2005). The methodological and empirical

contribution of this paper lies in the use of an in-depth longitudinal case study that

enables us to show the role of strategic alliance networks in the process of strategic

repositioning. Such a case study is often proposed to be the most appropriate research

strategy for studying networks aimed at new product development (Freeman, 1991). In

this sense we open up the black box of technology alliances. This provides us with new

insights that can add to current knowledge about technology alliances, which is often

primarily based on large-scale data-driven studies on the effect of technology alliances

on innovative performance (see e.g. Anand and Khanna, 2000; Duysters and

Hagedoorn, 2000; Vanhaverbeke et al., 2002).

The paper shows that exploration strategies, aimed at innovating and business

development, and exploitation strategies, which are primarily directed at making the

most of existing competences, require different network structures. Network behaviour

aimed at exploitation is compared to network behaviour aimed at exploration in the case

of IBM. Our main argument is that the networks needed to achieve these two particular

objectives differ substantially from each other and that IBM has consciously used a

carefully crafted networking strategy to support its strategic change project. This paper

shows how strategic change and network strategy can be linked effectively.

After assessing the connection between exploitation, exploration and networks,

IBM’s strategic change programme is discussed. Next IBM’s alliance network before,

during and after the strategic change programme is analysed and compared with IBM’s

peer group. The final section lists the implications and conclusions of our study.



4

2. Strategic change: exploration, exploitation and networks

Strategic change is often considered as a necessity for companies to survive in a

turbulent environment (Hamel and Prahalad, 1994). Intense international competition

and rapid technological change are often mentioned as primary motives for companies

to adapt their corporate strategy (Christensen, 1998; Eisenhardt and Tabrizi, 1995;

Sadowski et al., 2003). One way of facilitating strategic change is to engage in alliances

for the exploration of new capabilities and at the same time using alliances for the

exploitation of the existing knowledge base (March, 1991). Alliances are generally seen

as increasingly important instruments for learning (Kale et al., 2000; Khanna et al.,

1998). The process of learning in alliances basically boils down to the exchange of

technological knowledge and capabilities.

In order to come up with a more refined view on learning, we will make use of

March’s seminal distinction between exploitative and explorative learning. Exploitation

is generally associated with the refinement and extension of existing technologies.

Exploration on the other hand deals with experimentation with new alternatives and the

exploration of a new (technological) field. In this paper, we argue that there are

important differences between the two types of learning (see also March, 1991;

Chesbrough, 2003), which considerably affect the way in which companies should

make use of their external technology networks. Though March (1991) provides

examples of exploration and exploitation, he does not provide an operational definition

that can be used for empirical research in the field of alliances. In this paper, we will

distinguish between exploration and exploitation alliances by the type of R&D

activities, type of alliance and the characteristics of the partners involved.



5

As suggested above, exploration and exploitation strategies are not just internal to

the firm. Alliance networks are often used to support these strategies. In spite of the vast

body of literature on strategic technology alliances, only recently have authors started to

focus on the use of networks for exploitative or explorative learning (see e.g. Ahuja and

Lampert, 2001; Hagedoorn and Duysters, 2002; Rowley et al., 2000). Most, however,

have not taken a longitudinal view on strategic change by means of networking as

proposed by Freeman (1991). In addition, there has been very little attention to the

question of whether and how companies might transform their entire network from a

focus on exploitation towards a focus on exploration. Our view therefore extends

strategic management theory in terms of including alliance networks as a major

repositioning tool.

The existing contributions mainly argue that firms pursuing a strategy of exploration

for product development are most likely to establish alliances that are characterized by

‘weak ties’ (Granovetter, 1973). ‘Weak ties’ in this context imply that companies

exhibit low commitment to their alliances and team-up with non-familiar partners.

When exploring a particular new technology, companies may not want to enter into

inflexible forms of alliances, because they do not know whether the technology will

prove to be useful to them. In this case they value the opportunity to abandon the

alliance at any given moment (Duysters and De Man, 2003). Exploration alliances are

generally used when R&D efforts are aimed at the creation of radically new

technologies and new procedures (Gilsing and Nooteboom, 2006; Argyris and Schon,

1996).

Strong ties, characterized by intimate, recurrent and trustful relationships, on the

other hand are generally considered to be useful when firms aim at an exploitation



6

strategy (Krackhardt, 1992). In order to exploit knowledge and to make the most of

established technologies and products, trustworthy and intensive relationships with

partners are a prerequisite. Exploitation requires intensive knowledge exchange and the

creation of economies of scale. Both can be achieved in strong ties but to a much lesser

extent in weak ties, because only strong ties have the requisite intensity. Hence,

exploration strategies require lower-commitment R&D alliances in new technological

capabilities, since the focus is on learning new ideas from new partners. Exploitation

strategies on the other hand will benefit from high-commitment alliances in existing

technological capabilities (Koza and Lewin, 1998). In the literature we find some

scattered empirical evidence on this matter. Hansen et al. (2001), Afuah (2000) and

Rowley et al. (2000) find strong evidence that the value of strong and weak ties depends

on the type of learning and the external environment. Rowley et al. (2000) show that

strong ties are particularly effective for exploitation purposes and less effective for

exploration. The need for weak ties has been shown to be particularly high under

conditions of rapid technological change where the need for explorative learning is

highest (Afuah, 2000). In contrast to double-loop learning in exploration alliances,

exploitation alliances facilitate single-loop learning because they concentrate on current

technologies, process improvements and adapting existing procedures (Gilsing and

Nooteboom, 2006; Argyris and Schon, 1996).

-- Insert Table 1 --

Empirically, exploration networks differ from exploitation networks in three observable

ways (see Table 1). First, exploration networks will make use of flexible legal



7

organizational structures, whereas exploitation alliances are associated with legal

structures that enable long-term collaboration. In spite of their stability and the

advantages associated with higher levels of commitments in equity agreements, those

ventures seem to be less well equipped to deal with strategies of exploration (Duysters

and Hagedoorn, 2000; Spekman and Isabella, 2000). In terms of exploration, the

flexibility, speed and learning opportunities associated with non-equity agreements by

far outweigh the benefits associated with stability and improved commitment.

Exploration networks therefore have a preference for non-equity alliances, whereas

exploitation networks can be expected to exhibit a larger proportion of equity alliances

(Koza and Lewin, 1998). Joint development agreements and joint research pacts are

non-equity agreements with lower levels of commitment. Agreements with a high level

of commitment are equity-based relations like joint ventures. A measure for the extent

to which a network shifts towards an exploration strategy is obtained by counting these

alliance types across different time periods.

Second, in exploration networks partner turnover will be higher than in exploitation

alliances. A number of scholars (Ahuja, 2000; Burt 1998, 2000; Hansen, 1999; Hansen

et al., 2001) have suggested that different contingencies require different network

structures. Companies pursuing exploration strategies will change their partners more

often. This is in line with the findings of Granovetter (1973) and Khanna et al. (1998).

Comparing the composition of a network in period t with the network composition in

t+1 will show that the proportion of new partners in an exploration network is higher

than the proportion of new partners in an exploitation network. Exploration requires

access to a diversity of knowledge and a continuous scanning of new technological

opportunities. As these opportunities often arise outside existing partners, partner



8

turnover will be high. Exploitation requires intense collaboration. This takes time to

build up and benefits will accrue only after long-term collaboration. Consequently,

exploitation networks will have a higher proportion of the same partners over time than

exploration networks (Kale et al., 2000; Rowley et al., 2000).

A third measurable characteristic of networks that differs between exploitation and

exploration strategies relates to partner capabilities. In exploration networks companies

will look for partners with capabilities outside their own existing business. In

exploitation networks companies will tend to look for companies with similar

technological knowledge. Exploration strategies should lead to an innovation network

consisting of partners in new technological areas. Exploitation strategies on the other

hand may be expected to lead to an innovation network of partners in similar

technological areas. From an exploration perspective similarity among alliance partners

will decrease potential learning effects (Mowery et al., 1996; Duysters and Lemmens,

2003). The literature emphasizes that too much focus on local search might aid

exploitation but can lead firms to develop core rigidities (Leonard-Barton, 1995) and

subsequently can cause firms to fall into competency traps that decrease the amount of

explorative learning (Levitt and March, 1988). This is similar to what Benner and

Tushman (2002) and Ahuja and Lampert (2001) refer to as local versus distant search,

i.e. for capabilities close to or distant from the focal firm’s current skills and

capabilities.

We will attempt to take the discussion of exploration and exploitation to a higher

level by including a longitudinal view on companies. We are in particular interested in

showing the use of alliance networks for implementing strategic change processes. To

find out whether companies actually use and adapt their network when they are striving



9

for strategic change, we will study various network measures in the particular case of

IBM. In the 1990s IBM went through a major reorganization, changing from an

exploitation strategy towards an exploration strategy. This makes IBM a particularly

interesting case. We will observe whether IBM changed its network accordingly to

facilitate this change, using the network measures as stated in Table 1.

3. Strategic change at IBM: entering the internet era

The general outlines of the history of IBM are well known (Gerstner, 2002; IBM,

2002a). The relevant part of IBM's history for our research question starts in the 1980s

when IBM was the leading computer manufacturer in the world. At that time, it was

probably the best example of a vertically integrated corporation: almost all stages of

design, production and commercialization of computers remained internal to the firm

(Ernst, 2003; Lloyd and Phillips, 1994; Tushman and O’Reilly, 1996). This was the case

for semiconductors, hardware, operating systems, application software, and sales and

distribution. IBM was the world leader in computer manufacturing and it seemed that

the company’s leadership position would remain unchallenged for many years to come.

Because the position of IBM was seemingly comfortable, the company had not

developed sophisticated strategies to cope with fierce competition. This was also

evident from the organizational culture, which was focused on internal procedures

rather than concerned with the reality of a changing market environment (Hemp and

Stewart, 2004; Lloyd and Phillips, 1994; Tushman and O’Reilly, 1996). With the advent

of the ‘next big thing’, which was the rise of UNIX rather than personal computing,

IBM was suddenly under serious attack (Hamel, 2000; Meyer et al., 2005a). UNIX was

an ‘open’ operating system, supported by Sun and Hewlett-Packard, which offered



10

customers the first attractive alternative to IBM’s mainframe computers (Gerstner,

2002). In addition, IBM failed to see that personal computers (PCs) would be widely

used by business and enterprises, so the PC market was not a high priority for IBM. PCs

were not thought of as a major challenge to IBM’s core business of enterprise-wide

computing and systems integration (Balgobin and Pandit, 2001; Hamel, 2000). IBM

gave control over the operating system to Microsoft and the microprocessor to Intel, and

in the early nineties IBM’s leadership position started to deteriorate (Balgobin and

Pandit 2001; Hamel, 2000; Langlois and Robertson, 1992; Takahashi and Namiki,

2003). Fujitsu, Digital Equipment and Compaq were the competitors for hardware

components and were catching up fast (Curry and Kenney, 1999; Hamel, 2000). EDS

and Andersen Consulting were gaining ground in information services, while Intel and

Microsoft were more profitable in the PC market than IBM at that time (Hamel, 2000).

The once so comfortable position of ‘Big Blue’ was fading away at a very rapid pace. In

April 1993, Louis Gerstner was appointed as CEO and he had the difficult task to

transform IBM in such a way that it could regain its competitive position.

Two forces that emerged in the computer industry in the early 1990s were decisive for

IBM’s strategic change (Gerstner, 2002). The first was system integration, which

originated from customer demand. Customers increasingly valued companies that could

provide so-called integrated solutions: systems integration for business enterprises and

bundled software and services for the consumer market (Curry and Kenney, 1999;

Ghandour et al., 2004; Hemp and Stewart, 2004; Meyer et al., 2005b). Because of this

customer-driven force, IBM realized that in time the ICT industry would be service-led

rather than technology-led. The second force was the emergence of the networked

model of computing that would replace the stand-alone PCs that dominated the market



11

at the beginning of the 1990s (Curry and Kenney, 1999; Sweeney, 1998). The PC would

be one of the networking devices, but the management of the enormous flows of digital

information would be done on large-scale systems rather than desktop computers

(Gerstner, 2002). Thus, computing infrastructure and software would have the future. It

was clear that IBM had to turn its attention toward services and software.

Hamel (2000) provides an illustration of the strategic change that helped IBM transform

from a hardware manufacturer into a dominant service provider. A few visionary

individuals at IBM, who discovered the Internet as a potential source of future revenues,

brought about the major change that took place in the mid-nineties. This small group of

believers developed an Internet strategy for the corporation as a whole.

Palmisano, Gerstner’s successor, took e-business even further, using a different

strategy (Hemp and Stewart, 2004). He successfully made IBM’s organization flatter

and more flexible, which enabled the company to adapt more quickly and more

adequately to a rapidly changing competitive environment (BusinessWeek, 2003; Hemp

and Stewart, 2004). Palmisano’s strategy was to exit commodity hardware technologies

and concentrate on higher-margin software and services, and especially integrated (e-

business) solutions (BusinessWeek, 2004a; Ghandour et al., 2004; Hemp and Stewart,

2004).

In short, before Gerstner initiated its strategy change, IBM followed an exploitation

strategy. This is apparent from the fact that there was a strong focus on the internal

development of hardware and software in existing expertise areas of mainframe

computers and PowerPCs. IBM’s organizational culture was focused inwardly, where a

preoccupation with internal procedures hindered the company in understanding the

reality of drastically changing markets (Tushman and O’Reilly, 1996). After Gerstner’s



12

strategy change, IBM followed an exploration strategy. This is evident from the fact that

IBM was willing to sell and share technology in exchange for know-how in new areas

of software and service development (Gerstner, 2002; Takahashi and Namiki, 2003).

Whether this change also meant a change in IBM’s alliance network will now be

researched.

4. Data and sample

Whether IBM’s alliance network was redirected to support its move from an

exploitation strategy towards an exploration strategy is studied by means of looking at

the network composition in three pairs of years1 over five-year intervals: 1991/1992,

1996/1997, 2001/2002. The predictions based on Table 1 are that, over time, IBM starts

to make relatively more use of non-equity alliances, increases the number of new

partners in its network and looks for an increasing number of partners outside its

1991/1992 core business. Especially when it suffices to focus on an isolated aspect of a

case and it is not necessary to study cases in their entirety, repeated observation is a

good method to research the changes in relevant variables over time (Yin, 1989). In this

case, the IBM alliance portfolio is such an isolated aspect.

For our specific research question the alliance portfolio can be researched in

isolation from the multitude of other characteristics of IBM's transformation, and hence

repeated observation is applicable here. The longitudinal case study approach that we

use in this study is also based on event analysis, by singling out a period of strategic

change and taking snapshots of the IBM network before, during and after that period in

order to investigate the fit between the event (strategy change) and the company’s

1 Pairs of years are used, rather than individual years, in order to obtain a reasonable number of alliances for our study.



13

alliance network. This method of repeated observation around an event aims to increase

our understanding of network change, by combining qualitative with quantitative

analysis. The IBM case is selected because of the possibility to replicate our theoretical

predictions. In addition, a comparison is made with an IBM peer group to show that

IBM’s renewal process was markedly different. In this way, emerging theory about the

use of alliance networks for strategic repositioning can be extended (Eisenhardt, 1989).

In this paper, we argue that new strategic directions induce firms to adapt their alliance

networks in order to facilitate the required strategic change. Therefore, the need for

strategic change comes first, then firms adapt their alliance networks accordingly to

facilitate this change, and as a result the strategic re-positioning process in the market

can be concluded.

In addition to establishing a theoretical fit with the framework, the conclusions

about the use of networks for strategic change may also be in line with related research

into networks, as discussed in relation to Table 1. When this form of literal replication

exists, it would indicate a wider validity of the findings than just the IBM case. Our aim

here is to refine theory on alliance networks by studying the mechanisms companies use

to realize a strategic change from exploitation to exploration with the help of alliances.

We will test our theoretical predictions using alliance data from the MERIT-CATI

database and the CGCP database, which tracks alliance activity over the period covering

our study.2 The databases register the partners, their industry and the type of alliances

and thus contain all relevant information. The periods are chosen because they represent

three time periods relevant for IBM’s strategic change. In 1991/1992 IBM had an

2 More information on the CATI database can be found in Hagedoorn and Schakenraad (1992). The CGCP database is used as a complementary database for the period 2001-2002, for which the data were not available in the MERIT-CATI database. The data collection method of the CGCP database is similar to the MERIT-CATI database. For a description of the CGCP database, see http://www.cgcpmaps.com.



14

exploitation strategy. In 2001/2002 the effects of the strategic change towards an

exploration strategy should have materialized. The years 1996/1997 are years in which

the first effects of the strategic change initiated by Gerstner should have become visible.

Gerstner started in 1993 as CEO of IBM and he took some time to formulate the new

strategy. Next, it takes about one or two years before an alliance is set up, negotiated

and announced, so that 1996 will be the first year in which the strategic change could be

discernible in IBM’s alliance network. In addition, we compare alliance activities of

IBM with the alliance activities of the company’s main five competitors: Apple,

Compaq, DEC, Hewlett-Packard and Toshiba. The alliances of these firms are used as

benchmarks to determine to what extent the changes in the alliance network of IBM

differed from other main players in the same industry.

Alliance agreements can take a variety of forms, with a wide range in the level of

commitment. Alliance agreements include (cross-)licensing agreements, customer-

supplier contracts and standard-setting agreements. Since this is a study of innovation

networks, the alliances that are of particular relevance are those that involve joint

research and development or other forms of technology sharing (Koza and Lewin,

1998). We therefore study joint development agreements, joint research pacts and

research joint ventures.

IBM is historically by far the most active in alliance formation among its peer

group. The company has had more alliances than any other company in the ICT domain:

384 registered technology alliances with 227 different partners in the period from 1985

to 2002. These numbers make IBM an interesting case for further investigation.



15

5. IBM's alliance network: 1991-2002

In this section, we will investigate how networking strategies facilitated the business

transformation of IBM initiated by Gerstner. In order to do this, the networking

strategies of the early nineties, before Gerstner’s arrival in 1993, are compared with the

networking strategies in the mid-nineties and the beginning of the 21st century.

Most activities that characterized IBM in the late eighties and early nineties have been

gradually farmed out to multiple layers of specialized suppliers, giving rise to rapid

market segmentation and an ever-finer specialization. This has given rise to the co-

existence of complex, globally organized product-specific value chains (e.g. for

microprocessors, memories, board assembly, PCs, networking equipment, operating

systems, applications software, and sales & distribution). An important initial catalyst of

vertical specialization was the availability of standard components, which allowed for a

change in computer design away from the centralized IBM mainframe to decentralized

architectures, PC and PC-related networks (see e.g. Langlois and Robertson, 1992).

In order to see the significance of the overall corporate strategy change at IBM, the

network embeddedness of IBM in the period before Gerstner’s appointment is

compared with that in the ensuing period.

5.1 IBM's alliance network: 1991-1992

In the period 1991-1992 IBM engaged in 55 alliances, 42 of which were joint

development agreements and joint research pacts, nine were joint ventures and research

consortia and two were cross-licensing agreements (see Figure 1). From those 55

strategic alliances, 23 were in the field of computer manufacturing, mainly in the

development of microprocessors, and 23 in the field of software development, mostly



16

related to operating systems and software architecture. IBM had two important alliances

with Microsoft and Intel. Microsoft and IBM cross-licensed Windows New Technology

and with Intel, IBM had a long-term agreement on the development of microprocessors.

This latter agreement had a time horizon of 11 years, but was terminated in 1993. The

two agreements confirm IBM’s choice for the so-called Wintel personal computers.3

IBM has also been active in developing software for third parties. Examples of these

types of software development agreements are alliances with airline-companies such as

a flight reservation system for American Airlines (AMAIRL) in 1992. IBM is also very

active in developing communication networks, such as local area networks (Soh and

Roberts, 2003). The company is involved, to a lesser extent, in industrial automation,

such as CAD/CAM applications.

-- Insert Figure 1--

IBM intensively collaborates with personal computer and software developer Apple in

many fields of alliances within the ICT domain. The collaboration with Apple at first

sight seems to be a little odd, since IBM and Apple support different and competing

basic designs of computing (Hagedoorn et al., 2001). However, in the period 1991-1992

IBM and Apple shared ten strategic alliances, mainly related to the development of

microprocessors and software architecture. The technology developed in these alliances

is mainly related to microprocessors for PowerPCs and mainframe computers, and the

development of associated software, including network software, operating systems and

some multimedia applications. These technologies focus on RISC architecture, which

3 Wintel is a contraction of Windows and Intel. IBM had set the standard for the PC, Microsoft for the operating system and Intel for the microprocessor (see also Takahashi and Namiki, 2003).



17

both companies had been using for a number of years. Though non-equity R&D

agreements may indicate a strategy towards the exploration of new capabilities (Koza

and Lewin, 1998), it is clear that IBM and Apple were highly committed (Gulati, 1999;

Nooteboom, 1999; Uzzi, 1996, 1997; Walker et al., 1997). According to Granovetter the

strength of a tie is “a combination of the amount of time, the emotional intensity, the

intimacy, and the reciprocal services which characterize the tie” (Granovetter, 1973, p.

1361). All these factors are likely to be high in the case of the Apple-IBM alliance. The

long-term relationship between these two companies therefore indicates a strong tie for

exploiting technological capabilities in RISC architecture and software. Strategic

alliances between IBM and Apple are not reported in the period 1997-2002.


In the period 1996-1997, IBM had 32 strategic alliances, 27 of which were joint

development agreements, two were joint ventures, two cross-licensing agreements and

one a standardization agreement. What stands out immediately when comparing this

period (Figure 2) with the previous one (Figure 1) is the collaboration with multiple

partners and the increased complexity of the network configuration. In the period 1991-

1992 most agreements were bilateral, but in the period 1996-1997 there are some large

consortia involving many different partners. The multiple partnerships with Toshiba and

Motorola are all in developing microchips, one of IBM’s core competencies, indicating

that IBM was exploiting existing capabilities (March, 1991; Walker et al., 1997).

However, more and more of the partnerships in the period 1996-1997 were joint

development agreements in relatively new areas of expertise, indicating that exploration

of new technological capabilities was becoming more important to IBM (Khanna et al.,



18

1998; Koza and Lewin, 1998; March, 1991). This was especially true for joint R&D in

the development of multimedia and browser software. Before the mid 1990s, IBM had

no presence in the consumer software markets, but the company did have an increasing

number of alliances from 1991 onwards for developing products related to multimedia

and browser software. Along with these relatively new fields, existing technological

capabilities in microelectronics and computing were used for developing microchips

and improving PowerPC technology.


From 1996 onwards it was becoming clear that the pioneering work of the Internet

group at IBM (Hamel, 2000), discussed above, was being mutually reinforced by

corporate networking strategies. Before 1996, IBM had no alliance agreement related to

the development of Internet-related products or services. In 1996-1997 only 6 out of 32

alliance agreements dealt with the Internet, one of which was a joint venture with

Netscape, Oracle, Sony, Nintendo, Sega Enterprise and NEC. This joint venture was set

up for the development of Internet browsing software. From 1996 onwards, Internet-

related products and services were becoming more and more important. IBM has

gradually employed networking strategies in a variety of Internet-related products and

services based on the products developed by the Internet group, such as Internet

browsers, ThinkPad, WebSphere and other e-business applications. IBM’s core

competencies were still in computer hardware and software, but the focus of attention

was shifting to Internet and e-business solutions, a field unknown to the company before

1996. The relatively strong market position in e-business solutions that IBM established



19

around 2000 (see Figure 4) followed from a targeted alliance strategy with partners in

telecom and ICT consultancy, initiated in the early 1990s, such as the alliance with

Televerket on electronic data interchange and business consulting. Developing these

new capabilities clearly indicates an exploration strategy (Khanna et al., 1998; Koza and

Lewin, 1998; March, 1991). So, in a relatively short period of time IBM managed to

change from a hardware manufacturer into a global service provider. This was achieved

first through internal organizational transformation and second through a new portfolio

of networking strategies.


When comparing the network of alliances of 1996-1997 and 2001-2002, it appears that

alliances in computer manufacturing were becoming less important. In 1996, hardware

manufacturers like Motorola and Sun Microsystems were prominent partners of IBM

(Figure 2), whereas there are no alliances with these companies in 2001-2002 (Figure

3). On the other hand, alliances in software and telecommunications became more

important. The sharp increase in telecom and Internet alliances in the period 1996-1997

led to steady revenue growth in global services in the period 1996-2002 as compared to

revenues from software development and hardware manufacturing (Figure 4). Alliances

with Microsoft, Peoplesoft, and Citrix Systems in 2001-2002 (Figure 3), for instance,

show that IBM engaged in a wide variety of software development projects. In the field

of telecommunications, IBM was developing new products with network developers

Cisco and Nortel Networks, and leading mobile phone manufactures such as Ericsson,

Nokia and NTT DoCoMo in 2001-2002. These collaborations made IBM a strong

partner in telecommunications in North America, Europe and Asia.



20


The data presented in Figures 1-3 show that IBM has engaged mainly in alliances in the

fields of computer manufacturing and software development, the company’s traditional

core competencies, which might lead to the conclusion that IBM has mainly exploited

existing capabilities. Frequent partnerships in computer and software development

demonstrate IBM’s high level of commitment (Gulati, 1999; Krackhardt, 1992;

Nooteboom, 1999; Uzzi, 1996, 1997; Walker et al., 1997), in that the company

maintains relationships with a select group of partners in a wide variety of projects over

a longer period of time. However, especially in the period 1996-1997 and 2001-2002,

IBM established more relationships outside its existing technological capabilities. Thus,

gradually exploration of new capabilities was becoming more important (Khanna et al.,

1998; Koza and Lewin, 1998; March, 1991).

5.4 Alliance networks as a means for strategic repositioning

The revenue streams as shown in Figure 4 confirm the trend that IBM is exploring new

capabilities, whereby the company gradually transforms into a service-led company.

The products developed by the Internet group led to an increase of e-business activities

of IBM and eventually transformed the company into a global service provider. Not

surprisingly, hardware had been the core business activity in the second half of the

1990s, but in 2001 the revenue stream of global services outgrew the revenues of the

hardware division. The strong revenue growth in services was preceded by a strong

increase in alliances in telecom, software and Internet, and a decrease in alliance

activities in hardware manufacturing from 1996-1997 onwards. The global services



21

division generated revenues of almost $35 billion in 2001 compared to $30.5 billion for

hardware. The gap between the revenues of global services and hardware had grown

even further by 2002, at respectively over $ 36.4 billion and $ 27.5 billion (see Figure

4). Under Palmisano’s leadership the decreasing importance of the hardware division is

also evident from IBM’s recent deal to sell most of its PC division to the Chinese

company Lenovo (BusinessWeek, 2004b). Furthermore, IBM bought

PriceWaterhouseCooper’s consulting division in 2004, which is also a strong indication

of the growing importance of IBM's service business (InformationWeek, 2004).


In terms specifically of alliance type, speed of change of partners and the type of

capabilities sourced via alliances, Tables 2-4 show numerically what changes in the

IBM network took place. In addition, we can compare IBM’s alliances with those of the

company’s main five competitors in that period: Apple, Compaq, DEC, Hewlett-

Packard and Toshiba. These companies together dominated the PC market from the

early nineties till the late nineties. The numbers of these five companies have been

aggregated to make a comparison between IBM and one benchmark (Top-5) possible,

and to level out the potential volatility and inconsistencies of the figures of the five

individual companies.4 First, a change in networking strategies can be seen in a change

in the proportion of non-equity versus equity agreements. The first category of

agreements consists of joint development agreements and joint research pacts. These

non-equity joint research and development agreements indicate exploration strategies,

4 Note that DEC was acquired by Compaq in 1998 and Compaq merged with Hewlett-Packard in 2002.



22

as discussed in Koza and Lewin (1998). Equity agreements, such as joint ventures, are

usually associated with exploitation (Koza and Lewin, 1998). Though the pattern of

alliances is volatile, in terms of both the number of agreements and their type, it

becomes clear from Table 2 that equity agreements have become less important in

IBM’s networking strategies over the last decade. A relatively large proportion of high

commitment agreements and an abundance of licence agreements related to

microelectronics and computing characterized the period till 1992. After 1992, the

proportion of high commitment agreements decreased, while the proportion of lower

commitment R&D agreements increased. In the base period 1985-1990 and 1991-1992,

equity agreements took shares of respectively 12% and 18% of all alliance agreements.

In the period 1996-1997 and 2001-2002 these shares decreased to 6% and 5%,

respectively. When looking at the five main competitors of IBM (Top-5), we see a

similar pattern; in the period 1991-1992 on average 17% of alliances of the five

companies were based on equity alliances, while in 1996-1997 this proportion decreased

to 5% on average (see Table 2). In the period 2001-2002, however, the average

proportion of equity alliances increased again to 11%. Following Koza and Lewin

(1998), this trend seems to show that non-equity alliances, which indicate a tendency

towards exploration, are becoming slightly more important than equity alliances

(exploitation) after 1991-1992 for IBM. This also holds when comparing IBM with its

competitors, but the shift from non-equity to equity alliances is relatively small and

volatile. Therefore, no strong conclusions on a trend towards exploration can be made

based on the data on equity and non-equity alliances.

-- Insert Table 2--



23

Comparing the types of alliance partners in the three innovation networks with each

other and with the period 1985-1990 demonstrates that the alliance activities are

dominated by a search for new partnerships (Granovetter, 1973). Table 3 shows that the

share of new partners in the network compared to each of the previous periods is

relatively high. In the period 1991-1992, 63% of the partners were new to IBM, and in

the period 1996-1997, the share of new partners was 68%. The most notable change,

however, can be seen in the transition to the period 2001-2002, where the share of new

partners increased to 78%. This means that IBM collaborated earlier with only 13 out of

59 partners and all other partners were new to the firm compared to the period before

2001. When we compare IBM’s figures with those of its main competitors (Top-5), we

can conclude that IBM was renewing its partner network much faster than the other five

companies. In the period 1991-1992, on average only 38% of the partners of the five

main competitors were new in comparison to the period 1985-1990, compared to 63%

in IBM’s network (see Table 3). The average proportion of new partners in the

competitors’ networks increased in the periods 1996-1997 (59%) and 2001-2002 (66%),

but these proportions of new partners are much lower than in the case of IBM’s network

in the same periods (68% and 78% respectively).

Moreover, most of IBM’s partners in 2001-2002 come from a different

technological field than IBM’s. Only a few of the frequent partners of the periods 1991-

1992 and 1996-1997 in hardware manufacturing can be found in Figure 3. Instead,

IBM’s partners in the period 2001-2002 work mainly on client-based software and

services. This trend indicates that IBM is becoming more of a service-led company,



24

which again reflects Gerstner’s strategy of business transformation towards exploration

of new capabilities.

--Insert Table 3--

To get a better grip on the technological capabilities that are searched for, all alliances

that IBM engaged in are subdivided into six categories: computers, software, Internet,

telecommunications, microelectronics and other ICT related activities (Figure 5).5 The

focus of attention in strategic technology alliances of IBM over the years is rather

volatile, but alliances in computer and microelectronics manufacturing, and the

development of standard and dedicated software, are most common in the period 1985-

1992. In the period 1985-1990, half of IBM’s agreements were in computer

manufacturing and one-quarter in software development. In the period 1991-1992,

computer manufacturing and software development have equal shares of 42% and

naturally there were no alliances in Internet applications (see Figure 5). The five main

competitors (Top-5) show similar figures, 33% in software and 44% in hardware

alliances. However, strategic networking in the development of Internet applications

(both hardware and software) became more important after 1992. In 1996-1997, 19% of

IBM’s alliance agreements dealt with Internet applications in comparison to 10% of the

Top-5’s alliances. But especially in 2001-2002, the proportion of alliances in Internet

applications is remarkably high with a share of 55%. The five main competitors show

almost no growth in Internet alliances, with a small increase to 12%. Thus, IBM showed

a much greater increase in Internet alliances than its main competitors in the periods

5 These are categories that are common in the MERIT-CATI and CGCP database. A detailed description of all categories can be found in Duysters and Hagedoorn (1993).



25

1996-1997 and 2001-2002. In the light of March’s (1991) definition, this indicates that

IBM’s exploring strategies were more dominant in 2001-2002 than its exploitation

strategies.

--Insert Figure 5--

6. Implications and conclusions

In this paper we have described the role of alliance networks in strategic change

processes. We analyzed one of the most impressive strategic repositioning programmes

in the history of business, of how deliberate changes in the network structure enabled

IBM to fully reposition itself in the industry. The strategic change project that IBM

initiated in the 1990s is clearly facilitated by the changes brought about in its alliance

network. The network reflects the vast shift in the way IBM shaped its learning strategy

with network partners, changing it from exploitative learning towards explorative

learning.

The study of IBM’s network has a number of important implications. The first is

that company alliance networks can be used to facilitate strategic change inside a

company. This role of alliance networks has, so far, been relatively neglected in the

alliances literature. We show that strategic renewal is not just limited to an internal

project. The case of IBM shows that strategic repositioning can be facilitated by an

active use of the company network. Different strategies require different types of

networks. Learning strategies aiming at exploitation require different alliances and

alliance partners than learning strategies aimed at exploration. To be successful, new

company innovation strategies need to be translated into new networks (De Man, 2004).



26

We have shown that it is not the number of alliances which matters but the nature of the

alliance network which is most important. In fact, management attention and integration

costs may grow exponentially beyond a certain level of alliances (Duysters and De Man,

2003). In other words, a firm can start to suffer from information overload and

diseconomies of scale once it is involved in too many alliances at the same time. IBM

did well in that respect because it focused on changing the nature of its alliance

networks and did not focus on expanding the number of alliances as a goal in itself. This

paper, therefore, strongly suggests that the traditional view of large firms as inert

organizations that are unable to adapt swiftly to changing environmental conditions as

described in organizational ecology, evolutionary economic theory and in resource

based theories is no longer a valid perspective, given the opportunities that alliance

networks create for companies to reposition themselves in the market.

Second, this strategic change in the network can be brought about by two main

mechanisms: increasing the speed of change of partners and looking for partners in

areas outside existing competences. IBM has used both of these techniques to support

its new strategic direction. Even though some authors have claimed that networks

stimulate innovation (e.g. Chesbrough, 2003; Porter, 1990), the techniques that

companies employ for this have rarely been identified. The IBM case not only shows

these techniques, it also shows they are employed simultaneously to obtain the full

effect. Merely changing partners is not sufficient; the type of relationship with them

should change as well, as should the type of partner. From a practitioner’s perspective,

but also from an academic point of view, these findings are very important because,

apart from a few notable exceptions (Ahuja and Lampert, 2001; Hagedoorn and

Duysters, 2002; Rowley et al., 2000) the dominant stream of literature on alliances tends



27

to ignore a more in-depth view of various types of learning and the employment of

different alliance strategies to accommodate such learning. Somewhat surprisingly the

actual contribution of our associated variables (non-equity relations and low-

commitment alliances) is found to be rather low. This is interesting because the existing

literature seems to indicate otherwise. This may imply that in a situation of substantial

strategic change, high commitment to some partners is still important, for example

because knowledge exchange is easier in high commitment alliances. This helps a

company build up new capabilities in a new industry faster. In addition, finding reliable

allies in a new industry may be of such strategic importance that it is worth sacrificing

some flexibility in partner relationships. The thesis that low commitment alliances are

important for exploration may therefore be true only when companies remain active in

their own sector, but hold only partly when companies move swiftly into entirely new

sectors.

Third, this study suggests IBM has been consciously able to manage the shift in its

network. The alliance literature has paid some attention to change and network

dynamics. However most of the attention has been theoretical. Empirical tests of

network dynamics are virtually absent (De Man and Duysters, 2005). The vast majority

of large-scale studies into networks are limited to tests of networks at one point in time,

rather than across different time periods. Case studies have paid some attention to

dynamics, for example Gomes-Casseres (1996). Their focus however was not so much

on how companies adapt their alliance portfolios, but on changes in alliance

constellations on an industry or alliance group level. Dynamics at the level of a focal

firm is therefore an area that is not well-researched. There has been little investigation

into mechanisms that companies employ to create and sustain a dynamic alliance



28

portfolio, even though they are a relevant explanatory factor behind network dynamics,

as the IBM case shows. The IBM case supports the opinion that companies have a

choice in shaping their networks and are able to affect network composition. Of course

a powerful player like IBM may have more latitude in directing its network than smaller

companies, but clearly a deterministic view on networks is not supported by this case.

Rather, networks are shaped by the conscious actions of companies. These findings

definitely strengthen the call for a more dynamic approach to network research (Oliver,

2001).

A fourth conclusion pertains to the method used. By comparing snapshots of

networks over time in relation to a specific event, it was possible to reveal and

understand the main trends in the IBM network. This method is an extension of the

method of repeated observation in case studies (Yin, 1989). In this case the time period

for making observations is not chosen randomly, but connected to a specific event. The

combination of qualitative and quantitative analysis provides a more solid ground for

drawing conclusions than a purely qualitative description would have done.

Simultaneously it provides more understanding than a purely statistical network analysis

could have achieved, because it stays closer to real-life characteristics of alliances

(equity vs. non-equity, type of industry of partners, new partner or not). In the case of

IBM the method of repeated observations around an event works well: it delivers good

insights about relevant changes in IBM’s network related to the event of its change of

innovation strategy. The disadvantage of the method is that it may be vulnerable to the

choice of event and time-windows. Table 4 for example shows volatility in the number

of software and computer related alliances. This may be caused by the choice of year.

Another limit is that the focus has been on only one event. Even though this appears to



29

be the most important event for IBM over the period studied, other events occurring

during the period 1991-2001 may have had an impact on IBM’s network as well. The

method used does not allow us to evaluate the influence of those events. Hence, the

method is probably best applicable to research looking at the long-term developments of

networks around large-scale changes that require a number of years to elapse before

their full effect shows. More fine-grained case studies are required to obtain detail about

the impact of lesser events.

A question raised by this case relates to the effect on the network partners of

strategic change in a company. The alliance literature stresses the importance of

(personal) relationships, trust and reputation in alliances and networks (Gulati, 1995;

Jones et al., 1997). If these elements are relevant then restructuring a network could be a

painful and difficult affair. IBM however seems to have adapted its network fairly

quickly. Does this mean that these relational aspects are less important than is often

thought? Or do relational aspects apply mainly in a situation of exploitation, and if so,

what relational aspects are relevant for exploration networks? Or have the costs of

changing the network indeed been very high for IBM and its partners? The previous

study has not given answers to these questions. Surveying or interviewing partners may

help to shed light on this issue. This is especially relevant because little is known about

the relative importance of social ties versus business logic in networks. Both have been

recognized as relevant influences, but to what extent they strengthen or contradict each

other is unclear.

In this paper we have tried to improve our current thinking about strategic

repositioning efforts by showing the importance of strategic alliance networks; an

importance thus far neglected in the literature. The IBM case shows that strategic



30

change is facilitated by changing the company network and that companies are able to

change the focus of their network from exploitation towards exploration. The translation

of the strategic change project into a new alliance network was realized by changing the

type of alliance, changing the partners and bringing new capabilities into the network.

By describing the IBM case we hope to have proved that an integrated approach linking

strategic change to network change not only has theoretical but also managerial merit.

References

Afuah, A., 2000. How much do your co-opetitors’ capabilities matter in the face of

technological change? Strategic Management Journal 21(3), 387–404.

Ahuja, G., 2000. Collaboration networks, structural holes, and innovation: a

longitudinal study. Administrative Science Quarterly 45(3), 425-455.

Ahuja, G. and Lampert, C.M., 2001. Entrepreneurship in the large corporation: a

longitudinal study of how established firms create breakthrough inventions.

Strategic Management Journal 22(6), 521-543.

Anand, B.N., Khanna, T., 2000. Do firms learn to create value? The case of alliances.

Strategic Management Journal 21(3), 295-315.

Argyris, C., Schon, D.A., 1996. Organizational Learning II: Theory, Method and

Practice. Addison-Wesley, New York.

Balgodin, R., Pandit, N., 2001. Stages in the turnaround process: the case of IBM UK.

European Management Journal 19(3), 301-316.



31

Ballot, G., Taymaz. E., 1997. The dynamics of firms in a micro-to-macro model: the

role of training, learning and innovation. Journal of Evolutionary Economics 7, 435-

457.

Benner, M.J., Tushman, M., 2002. Process management and technological innovation: a

longitudinal study of the photography and paint industries. Administrative Science

Quarterly 47(4), 676-706.

Bouchikhi, H., Kimberly, J.R., 2003. Escaping the identity trap. MIT Sloan

Management Review 44(3), 20-26.

Burt, R.S., 1998. The network structure of social capital. Paper presented at the Social

Network and Social Capital Conference, Duke University, Durham, NC.

Burt, R.S., 2000. The network structure of social capital. Research in Organizational

Behavior 22, 345-423.

BusinessWeek, 2003. The new Blue: Lou Gerstner saved Big Blue. Now it’s up to new

CEO Sam Palmisano to restore it to greatness. 17 March 2003, 80-88.

BusinessWeek, 2004a. Sam Palisano’s many messages. Online edition, 6 December

2004.

BusinessWeek, 2004b. Big Blue’s big coup. 20 December 2004, 35-36.

Chesbrough, H.R., 2003. Open Innovation. Harvard Business School Press, Cambridge,

MA.

Christensen, C.M., 1998. Innovation and the General Manager. Irwin McGraw-Hill,

Boston, MA.

Curry, J., Kenney, M., 1999. Beating the clock: corporate responses to rapid change in

the PC industry. California Management Review 42(1), 8-36.



32

De Man, A.P., 2004. The Network Economy: Strategy, Structure and Management.

Edward Elgar, Aldershot.

De Man, A.P., Duysters, G., 2005. Collaboration and innovation: a review of the effects

of mergers, acquisitions and alliances on innovation. Technovation 25(12), 1377-

1387.

Duysters, G.M., De Man, A.P., 2003. Transitory alliances: an instrument for surviving

turbulent industries? R&D Management 33(1), 49-58.

Duysters, G.M., Hagedoorn, J., 1993. The Cooperative Agreements and Technology

Indicators (CATI) Information System. MERIT, Maastricht.

Duysters, G., Hagedoorn, J., 2000. A note on organizational modes of strategic

technology partnering. Journal of Scientific & Industrial Research 58(8-9), 640-649.

Duysters, G.M., Lemmens, C.E.A.V., 2003. Alliance group formation: enabling and

constraining effects of embeddedness and social capital in strategic technology

alliance networks. International Studies of Management & Organization, 33(2), 49-

68.

Eisenhardt, K.M., 1989. Building theories from case study research. Administrative

Science Quarterly 14(4), 532-550.

Eisenhardt, K.M. and Tabrizi, B.N., 1995. Accelerating adaptive processes: product

innovation in the global computer industry. Administrative Science Quarterly 40(1),

84-110.

Ernst, D., 2003. Digital information systems and global flagship networks: how mobile

is knowledge in the global network economy? In: Christensen, J.F. and Maskell, P.

(Eds.), The Industrial Dynamics of the New Digital Economy. Edward Elgar

Publishers, Cheltenham, pp. 151-176.



33

Freeman, C., 1991. Networks of innovators: a synthesis of research issues. Research

Policy 20(5), 499-514.

Geroski, P.A., 2001. Exploring the niche overlaps between organizational ecology and

industrial economics. Industrial and Corporate Change, 10(2), 507-540.

Gerstner, L.V., 2002. Who says elephants can’t dance? Inside IBM’s historic

turnaround. Harper Collins Publishers, London.

Ghandour, F.A., Swartz, P., Grenek, H.M., Roberts, E.B., 2004. E-Business

transformation via alliance clusters. Technology Analysis & Strategic Management

16(4), 435-455.

Gilsing, V., Nooteboom, B., 2006. Exploration and exploitation in innovation systems:

the case of pharmaceutical biotechnology. Research Policy 35(1), 1-23.

Gomes–Casseres, B., 1996. The Alliance Revolution. Harvard University Press,

Cambridge, MA.

Granovetter, M., 1973. The strength of weak ties. American Journal of Sociology 78(6),

1360-1380.

Guillén, M.F., 2004. Structural inertia, imitation, and foreign expansion: South Korean

firms and business groups in China, 1987-95. Academy of Management Journal

45(3), 509-525.

Gulati, R., 1995. Does familiarity breed trust? The implications of repeated ties for

contractual choice in alliances. Academy of Management Journal 38(1), 85-112.

Gulati, R., 1999. Network location and learning: the influence of network resources and

firm capabilities on alliance formation. Strategic Management Journal 20(5), 397-

420.



34

Hagedoorn, J., 1993. Understanding the rationale of strategic technology partnering -

interorganizational modes of cooperation and sectoral differences. Strategic

Management Journal 14(5), 371-385.

Hagedoorn, J., Caryannis, E., Alexander, J., 2001. Strange bedfellows in the Personal

Computer industry: technology alliances between IBM and Apple. Research Policy

30(5), 837-849.

Hagedoorn, J., Duysters, G., 2002. Learning in dynamic inter-firm networks – the

efficacy of quasi-redundant contacts. Organization Studies 23(4), 525-548.

Hagedoorn, J., Sadowski, B.M., 1999. The transition from strategic technology alliances

to mergers and acquisitions: an exploratory study. Journal of Management Studies

36(1), 87-107.

Hagedoorn, J., Schakenraad, J., 1992. Leading companies and networks of strategic

alliances in information technologies. Research Policy 21(2), 163-190.

Hamel, G., 2000. Waking ip IBM. How a gang of unlikely rebels transformed Big Blue.

Harvard Business Review 78(4), 137-146.

Hamel, G., Prahalad, C.K., 1994. Competing for the Future. Harvard Business School

Press, Boston, MA.

Hannan, M., Freeman, J., 1984. Structural inertia and organisational change. American

Sociological Review 49, 149–164.

Hansen, M.T., 1999. The search-transfer problem: the role of weak ties in sharing

knowledge across organization subunits. Administrative Science Quarterly 44(1),

82-111.



35

Hansen, M.T., Podolny, J.M., Pfeffer, J., 2001. So many ties, so little time: a task

contingency perspective on corporate social capital. Research in the Sociology of

Organizations 13(3), 232 -248.

Hemp, P., Stewart, T. A., 2004. The HBR interview: Samuel J. Palmisano. leading

change when business is good. Harvard Business Review, 82(12), 60-70.

IBM, Annual Reports 1996-2002. IBM Corporation, Armonk, NY.

IBM, 2002a. A century of innovation. Flash presentation of IBM’s history, downloaded

from http://www.ibm.com in June 2002.

InformationWeek, 2004. Palmisano era in full swing at IBM. Online edition, 23 August

2004.

Johnson-Cramer, M.E., Cross, R.L., Yan, A., 2003. Sources of fidelity in purposive

organizational change: lessons from a re-engineering case. Journal of Management

Studies 40(7), 1837-1870.

Jones, C., Hesterly, W.S., Borgatti, S.P., 1997. A general theory of network governance:

exchange conditions and social mechanisms. Academy of Management Review

22(4), 911-945.

Kale, P., Singh, H., Perlmutter, H., 2000. Learning and protecting of proprietary assets

in strategic alliances: building relational capital. Strategic Management Journal

21(3), 217-237.

Khanna, T., Gulati, R., Nohria, N., 1998. The dynamics of learning alliances:

competition, cooperation, and relative scope. Strategic Management Journal 19(3),

193-210.

Koza, M.P. and Lewin, A.Y., 1998. The co-evolution of strategic alliances.

Organization Science 9(3), 255-264.



36

Krackhardt, D., 1992. The strength of strong ties: the importance of Philos in

organizations. In: Nohria, N. and Eccles, R.G., Networks and Organizations:

Structure, Form, and Action. Harvard Business School Press, Boston, MA, pp. 216-

239.

Langlois, R.N., Robertson, P.L., 1992. Networks and innovation in a modular system:

lessons from the microcomputer and stereo component industries. Research Policy

21(4): 297-313

Lavie, D., Rosenkopf, L., 2006. Balancing exploration and exploitation in alliance

formation. Academy of Management Journal 49(4), 797-818.

Leonard-Barton, D., 1995. Wellsprings of Knowledge. Harvard Business School Press,

Cambridge, MA.

Levitt, B., March, J.G., 1988. Organizational learning. Annual Review of Sociology 14,

319-340.

Lloyd, G., Phillips, M., 1994. Inside IBM: strategic management in a federation of

businesses. Long Range Planning 27(5), 52-63.

Louçã, F. and Mendonça, S., 2002. Steady change: the 200 largest manufacturing firms

throughout the 20th century. Industrial and Corporate Change 11(4), 817-845.

March, J.G., 1991. Exploration and exploitation in organizational learning. Organization

Science 2(1), 71-87.

Markides, C.C., P.J. Williamson, 1994. Related diversification, core competences and

corporate performance. Strategic Management Journal 15(Summer), 149-165.

Meyer, M.H., Anzani, M., Walsh, G., 2005a. Innovation and enterprise growth.

Research Technology Management 48(4), 34-44.



37

Meyer, M.H., Anzani, M., Walsh, G., 2005b. Organizational change for enterprise

growth. Research Technology Management 48(6), 48-56.

Mowery, D.C., Oxley, J.E., Silverman, B.S., 1996. Strategic alliances and interfirm

knowledge transfer. Strategic Management Journal 17, Special Issue: Winter, 77-91.

Narula, R., 2002. Innovation systems and ‘inertia’ in R&D location: Norwegian firms

and the role of systemic lock-in. Research Policy 31(5), 795-816.

Nooteboom, B., 1999. Innovation and inter-firm linkages: new implications for policy.

Research Policy 28(8), 793-805.

Oliver, A., 2001. Strategic alliances and the learning life-cycle of biotechnology firms.

Organization Studies 22(3), 467–489.

Owen-Smith, J., Powell, W.W., 2004. Knowledge networks as channels and conduits:

the effects of spillovers in the Boston biotechnology community. Organization

Science 15 (1), 5-21.

Porter, M.E., 1990. The Competitive Advantage of Nations. Macmillan, London.

Powell, W.W., Koput, K.W., Smith-Doerr, L., 1996. Interorganizational collaboration

and the locus of innovation: networks of learning in biotechnology. Administrative

Science Quarterly, 41(1): 116-145.

Rothmaerler, F.T., 2001. Complementary assets, strategic alliances, and the incumbent’s

advantage: an empirical study of industry and firm effects in the biopharmaceutical

industry. Research Policy 30(8), 1235-1251.

Rothmaerler, F.T., Deeds, D.L., 2004. Exploration and exploitation alliances in

biotechnology: a system of new product development. Strategic Management

Journal 25(3), 201-221.



38

Rowley, T., Behrens, D., Krackhardt, D., 2000. Redundant governance structures: an

analysis of structural and relational embeddedness in the steel and semiconductor

industries. Strategic Management Journal 21(3), 369-386.

Sadowski, B.M., Dittrich, K, Duysters, G.M., 2003. Collaborative strategies in the event

of technological discontinuities: the case of Nokia in the mobile telecommunication

industry. Small Business Economics 21(2), 173-186.

Soh, P.-K., Roberts, E.B., 2003. Networks of innovators: a longitudinal perspective.


Spekman, R.E., Isabella, L. 2000. Alliance Competence: Maximizing the Value of your

Partnerships. John Wiley & Sons, New York.

Suarez, F.F., Olivia, R., 2005. Environmental change and organizational transformation.

Industrial and Corporate Change 14(6), 1017-1041.

Sweeney, D.M., 1998. Global market trends in the networked era. Long Range Planning

31(5), 672-683.

Takahashi, T., Namiki, F., 2003. Three attempts at “de-Wintelization” Japan’s TRON

project, the US government’s suits against Wintel, and the entry of Java and Linux.


Tushman, M.L., O’Reilly III, C.A., 1996. Ambidextrous organizations: managing

evolutionary and revolutionary change. California Management Review 38(4), 8-30.

Uzzi, B., 1996. The sources and consequences of embeddedness for the economic

performance of organizations: the network effect. American Sociological Review

61(4), 674-698.

Uzzi, B., 1997. Social structure and competition in interfirm networks: the paradox of

embeddedness. Administrative Science Quarterly 42(1), 35-67.



39

Vanhaverbeke, W., Duysters, G., Noorderhaven, N., 2002. External technology

sourcing through alliances and acquisitions. an analysis of the ASIC industry.

Organization Science 13(6), 714-733.

Walker, G., Kogut, B., Shan, W., 1997. Social capital, structural holes and the formation

of an industry network. Organization Science 8(2), 109-125.

Yin, R.K., 1989. Case Study Research: Design and Methods. Sage, London.



40

Table 1: Network characteristics for exploration and exploitation strategies

Exploration Exploitation

Alliance type Non-equity alliances; few

equity alliances

Relatively high number of

equity alliances

Speed of changes of

partners

Higher: many new

partners enter the network

Lower: few new partners

enter the network

Type of partner

capabilities

Partners with different

technologies

Partners with similar

technologies in same

business



41

Table 2: Non-equity versus equity agreements

IBM TOP-5 Competitors

Number of

equity alliances

Number of non-

equity alliances

Avg. number of

equity alliances

Avg. number of non-

equity alliances

1991-1992 10 (18%) 45 (82 %) 3.6 (17%) 17.8 (83%)

1996-1997 2 (6%) 30 (94 %) 0.8 (5%) 14.8 (95%)

2001-2002 3 (5%) 57 (95 %) 4.8 (11%) 37.0 (89%)

Source: MERIT-CATI database (1985-1996); CGCP database (1997-2002)



42

Table 3: New partners in IBM’s innovation network in comparison with Top-5

IBM TOP-5 Competitors

Number

of partners

New

partners

% New

partners

Avg. number

of partners

Avg. number of

new partners

% New

partners

1991-1992 44 28 63% 19.8 7.8 38%

1996-1997 44 30 68% 17.2 9.6 56%

2001-2002 59 46 78% 35.5 23.0 65%




43

Figure 1: IBM’s innovation network 1991-1992

Source: MERIT-CATI database



44


Source: MERIT-CATI database (1996); CGCP database (1997)



45


Source: CGCP database



46

Figure 4: Revenues of IBM 1996-2002

0

5000

10000

15000

20000

25000

30000

35000

40000

1996 1997 1998 1999 2000 2001 2002 Year

Million $

Global Services Hardware Software

Global Financing Enterprise Investments/Other

Source: IBM (Annual Reports 1996 to 2002).



47

Figure 5: Search for capabilities in innovation networks of IBM and Top-5 competitors

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

IBM Top-5 IBM Top-5 IBM Top-5

1991-1992 1996-1997 2001-2002

OtherMicroTelecomInternetSoftwareComputers

55 106 32 63 60 135


Documents

Strategic repositioning by means of alliance …...Strategic repositioning by means of alliance networks: The case of IBM Koen Dittrich (corresponding author) RSM Erasmus University,