Contact information

Thomas Schmidt

Freie Universität Berlin

School of Business & Economics

DFG-Doctoral Program Research on Organizational Paths

Garystr. 21, 14915 Berlin, Germany

E-Mail: thomas.schmidt@fu-berlin.de

Timo Braun

Freie Universität Berlin

School of Business & Economics

Management Department

Chair for Inter-firm Cooperation

Boltzmannstr. 20, 14195 Berlin, Germany

E-Mail: timo.braun@fu-berlin.de

1

Submission to 3rd

International Conference on Path Dependence

February 17-18, 2014

Freie Universität Berlin

School of Business & Economics

When cospecialization leads to rigidity:

Why even SAP couldn’t unlock interorganizational path dependence

--- anonymized ---

2

When cospecialization leads to rigidity:

Why even SAP couldn’t unlock interorganizational path dependence

Abstract

By taking advantage of interorganizational asset complementarity (cospecialization), interfirm

relationships allow to gain relational advantages. Nonetheless, in the long run, firms are sometimes

found to be unable to adapt these interorganizational arrangements (in order to access new markets

with innovative products). We investigate this problem by conducting a qualitative case study of

SAP’s initiative to enter the SME market with a new product and partner strategy from 2007 until

2013. We contrast these findings with a historical analysis of the rise of SAP’s product and partner

strategy in the last four decades. This research design allows us to relate current rigidities to

historical roots. We find that organizational and interorganizational long-term effects of co-

specialization can become barriers for path-breaking initiatives. In particular, we identify two

specific pitfalls for change initiatives in vertically disintegrated value chains: boundedness and

asynchronicity.

Keywords: path dependence, cospecialization, complementary assets, interorganizational

relationships, alliances, complementarity effects, ERP, SAP

Introduction

Interorganizational relationships are rarely associated with inflexibility, inertia or path dependence.

Quite the contrary, networks and alliances are rather assumed to increase flexibility (Powell, 1990)

and adaptability (Gulati, Lawrence, & Puranam, 2005). This positive view on networks and

alliances was influenced to a large extent by the debate on core competencies among scholars and

practitioners (Prahalad & Hamel, 1990; Quinn & Hilmer, 1994). According to this way of thinking,

firms should focus on their core business to gain benefits from specialization and outsource all non-

core activities. While some non-core activities can easily be accessed by market transactions, there

are complementary or cospecialized assets outside of the boundaries of a firm that can only be

accessed by long-term cooperation strategies (Teece, 1992). A large body of literature highlights the

role of complementarities for interorganizational relationships (e.g. Dyer & Singh, 1998;

Hagedoorn, 1993).

Yet, just a few articles render the topic problematic and ask for negative long-term effects like

3

interorganizational path dependence. The literature has pointed occasionally to the problem of

intraorganizational effects caused by outsourcing – that increasing disintegration may promote

specialized “silos” – thus inhibiting systemic innovation (Chesbrough & Teece, 1996; Jacobides &

Winter, 2005). However, there is still a lack of empirical research that investigates the long-term

rigidities of cospecialization processes in interorganizational arrangements (for an exception see

Weigelt, 2009). The aim of this study is to show that – when strong self-reinforcing

interorganizational complementarity effects are at work – interorganizational processes of the

participating organizations could fall victim to strategic path dependence, thus even path breaking

initiatives fail (Sydow, Schreyögg, & Koch, 2009).

We investigate this phenomenon by conducting a qualitative multi-level case study of a failed

initiative of the world largest enterprise software vendor SAP to enter the SME market with a new

product and partner strategy from 2008 until 2013. We contrast these findings with a historical

analysis of the rise of SAP’s established product and partner strategy in the last four decades. This

research design allows us to relate current rigidities to historical roots. Thereby our study seeks to

provide answers to the following research questions:

Why do firms with successful interorganizational strategies fail to enter new markets?

What are the organizational and interorganizational long-term effects of cospecialization

processes between firms and how do these effects inhibit change?

The research problem is studied theoretically with an extension of the theory of organizational path

dependence (Sydow et al., 2009) for interorganizational settings – integrating the concept of

cospecialization (Teece, 1992). We find that the complementarity effects of cospecialization

processes enfold an “avalanche-like” (Sydow et al., 2009: 697) dynamic leading to relational

advantages. This positive feedback further stabilizes the interfirm cooperation pattern and shapes

the identity of the participating organizations. This stability becomes an issue against the backdrop

of environmental change.

Relational advantages are volatile due to innovations by competitors and changing market

environments. As managers become aware, they may make strong efforts to restructure the value

chain and to reintegrate and/or develop alternative interorganizational cooperation patterns. But

even though decision makers see the need for change, they are found to be unable to disrupt the

established pattern of cooperation. This seems puzzling, but we work out why it is well explained

by an extension of the theory of organizational path dependence. In particular, we show with our

study that the problem of path dependence might become even more severe, when cospecialized

activity systems span across the boundaries of different specialized firms. It is well known that

4

complementarity effects within organizations are not only a source of competitive advantage, but

can also be the roots of rigidities (Leonard-Barton, 1992) – especially when activity systems with

strong complementarity effects (Stieglitz & Heine, 2007) are confronted with radical innovations

(Henderson & Clark, 1990). But considering large partner networks that can consist of thousands of

organizations multiplies the problem with different cospecialization trajectories within each partner

firm. Thus, it is a challenging endeavor to restructure a whole value chain at once, because it can

imply, for example, a massive reintegration of capabilities. As we analyze a timeframe of four

decades within our case study, we are able to show how interorganizational cospecialization

processes evolve during a long time period. By this, we identify reasons why there are specific

pitfalls for change initiatives within interorganizational path-dependent settings.

Our study comprises three major contributions to organizational and strategy research: First, we

adapt the concept of path dependence to an interorganizational setting. In the past, single

organizations were in the focus of this research stream while studies on interfirm relationships were

scarce – yet these relationships are a critical feature of the business model of many corporations and

may be also subject to inertia or path dependence (Gulati, Nohria, & Zaheer, 2000). Second, our

theoretical framework integrates path dependence theory with the concept of cospecialization and

complementarity in the tradition of Teece (1992). Third, the explorative approach allows us to

identify conditions for the failure of path breaking initiatives. Based on our case study we show that

bounded change within organizational subunits and asynchronous change among the collaborating

partners are the major reasons for relapsing to old strategic paths. This has implications not only for

research but also for the (practical) management of interorganizational relationships.

In the next section we will introduce the characteristics of interorganizational path dependence and

review important studies on complementarity and cospecialization in alliances. Then we set out our

research methodology and present the empirical results of our explorative case study. Thereafter, we

show how the findings relate to our theoretical framework and how we can add to and refine

previous research. The concluding section summarizes and reflects on the limitations as well as the

contributions of this study.

Theory

From organizational to interorganizational path dependence

The concept of path dependence was developed by Brian Arthur (1989, 1994) and Paul David

(1985) and it explains why even inferior technologies can become dominant standards. Early events

like the early adoption of a technology by a significant user base might lead to “increasing returns”

5

(Arthur 1989, 1994) and under certain circumstances to a lock-in-situation. The concept is useful to

understand the role of history not only for the adoption of technologies, but also for other social

processes. On this basis, the theory of organizational path dependence (Sydow et al., 2009) provides

an explanation of how self-reinforcing mechanisms (phase II) lead an organization from

contingency (phase I) towards lock-in (phase III). As shown in Figure 1, the potential (strategic)

options of organizations are reduced continuously in the course of becoming path dependent (see

Figure 1).

Figure 1: The constitution of an organizational path (Sydow et al., 2009: 692)

The most prominent self-reinforcing mechanisms in path dependence research on organizations are

complementarity effects and coordination effects. Complementarities on the one hand are described

as the synergy surplus that is generated by the combination of at least two separated elements

(Ennen & Richter, 2010; Milgrom & Roberts, 1995; Porter & Siggelkow, 2008; Stieglitz & Heine,

2007). This advantage might lead to positive feedback and thereby to virtuous (or vicious) cycles

(Masuch, 1985). Coordination effects on the other hand “build on the benefits of rule-guided

behavior” (Sydow et al., 2009: 699). Well-established organizational rules coordinate the activities

of multiple actors and make unilateral deviation unattractive. Different studies show that

organizations that are stuck too tightly to their established rules fail to respond to environmental

change (Gilbert, 2005; Koch, 2008, 2011; Tripsas & Gavetti, 2000). Complementarity as well as

coordination are also key concepts in the literature on interorganizational relationships. In this

respect, there is some evidence for relational lock-ins (Maurer & Ebers, 2006), persistence in

partner selection (Inkpen & Ross, 2001; Li & Rowley, 2002; Patzelt & Shepherd, 2008) and

interorganizational path dependence (Burger & Sydow, 2012; Kuschinsky, 2008; Levering, Ligthart,

Noorderhaven, & Oerlemans, 2013; Mallach, 2012).

6

Cospecialization in interorganizational relationships

Resource complementarity is an important explanation for the emergence of interorganizational

relationships in general and for alliances and networks in particular (Dyer & Singh, 1998; Gulati,

1995; Hamel, Doz, & Prahalad, 1989; Nohria & Garcia-Pont, 1991). This is particularly true for

innovating firms that do not possess all cospecialized assets necessary to appropriate value from

innovation (Chung, Singh, & Lee, 2000; Colombo, Grilli, & Piva, 2006; Hess & Rothaermel, 2011;

Parmigiani & Mitchell, 2009; Rothaermel & Boeker, 2008; Rothaermel, 2001, 2001; Santoro &

McGill, 2005; Teece, 1992). Yet, the way the notion of complementarity is used in this context often

implies a simple surplus that comes from the combination of two or more separated elements, for

example through the interplay of a technological innovation with a specific marketing competence.

However, the original concept of cospecialized assets (Teece, 1986, 1992) is more subtle and also

differentiates between the directions of the complementary relationships. Accordingly, an asset is

specialized, if there is a unidirectional dependence on another asset and assets are cospecialized,

when the dependence is bidirectional.

A further elaboration on the concept of cospecialized assets was developed by Jacobides, Knudsen

and Augier (2006). They unbundle the implicit dimensions of the concept: complementarity and

mobility.

Figure 2: Complementarity vs. mobility: dependence and complementarities (Jacobides et al., 2006: 1207)

Thus, one feature of cospecialized assets is (a positive form of) complementarity in the sense of

„superior returns to a combination of two or more assets, i.e. complementarity in products, services,

and processes“ (Jacobides et al., 2006: 1206). The other dimension is mobility which can be seen as

the opposite of asset specifity, meaning that assets can be complemented not only by one, but by

many partner firms.

7

„The second is (bilateral) dependence in the sense of the number of assets that can potentially enter a

combination, with negligible switching costs, i.e. mobility in assets that are components of a

combination” (Jacobides et al., 2006: 1206).

Assets with high complementarity and high mobility are different from Teeceian cospecialized

assets. The Teeceian concept can be viewed as an issue of conjunctural causality (Mackie, 1965;

Ragin, 2000), meaning in the case of complete cospecialization that one cospecialized asset would

be insufficient (worthless) without the other and vice versa.

“With co-specialization, joint use is not only value enhancing; It also will be asset specific (i.e. the co-

specialized assets do not have a market in which they can be sold for their full value)” (Pitelis & Teece,

2010).

This is the reason, why Teece advocates to vertically integrate critical, cospecialized assets, when

competition is high and approprability regimes are weak. By contrast, Jacobides et al. (2006)

conceive a concept of cospecialization stating that some firms are able to benefit from asset

complementarity not only in one, but in many interorganizational relationships. These firms are

willing to give up a part of the value chain and transfer parts of the value creation process to partner

firms. Means of managing this kind of asset complementarity are, for instance, open standards. With

this conceptualization, the idea of cospecialization becomes linkable to the discussion on platforms,

ecosystems and keystone advantages (Gawer, 2011; Iansiti & Levien, 2004).

Such kind of managed interorganizational cospecialization is a frequent phenomenon in industries

with fast innovation cycles and technological complementarities – like the microelectronics or the

software industry (Hagedoorn, 1993) and especially in industries with high (indirect) network

effects. The availability of complementary goods and services is a key for example for software

vendors to gain installed base advantages (Schilling, 1999) and to exploit technological path

dependence (Arthur, 1989, 1994). In a similar vein Venkatraman, Lee & Iyer (2008) find evidence

that software firms need to “interconnect to win” (see also Lee, Venkatraman, Tanriverdi, & Iyer,

2010; Shapiro & Varian, 1999). On the contrary, in these markets with increasing returns also

promising products with a good basis value can fall victim to the penguin effect (Katz & Shapiro,

1985), a cautious behavior of customers and other market actors, when the critical installed base is

not achieved quickly enough.

Cooperation patterns in complementarity-based alliances

The idea of managed complementarity also opens the concept of cospecialization for a broader unit

of analysis. When the mobility of assets is high, cospecialization processes do not only take place

on the level of one specific dyadic relationship, but also on a broader level of an industry or on a

network level. A focal firm may increase the value of its focal asset by cooperating not only with

one, but with a whole network of partner firms – thereby leveraging complementarity effects.

8

However, this managed complementarity is demanding, it requires the patterning of the

collaboration processes, i.e. a pattern of task alignment in the value chain. Jacobides et al. (2006)

describe these patterns as “industry architectures”.

“Thus, industry architectures provide two templates, each comprising a set of rules: (1) a template

defining value creation and the division of labor, i.e. who can do what and (2) a template defining value

appropriation and the division of surplus, or revenue, i.e. who gets what” (Jacobides et al., 2006: 1205).

In other words, the exploitation of complementarity effects across organizational boundaries also

increases the necessity of coordination. The management literature on complementarities points to

the effort of coordination that grows with increasing complementarities: “If corporate resources are

complementary, the need for some kind of coordination is apparent, since the added value of one

resource depends on the use of other resources and their individual deployment has to be

consistent” (Stieglitz & Heine, 2007: 3).

The same seems to be true for the coordination of interorganizational relationships (Gulati & Singh,

1998; Van de Ven & Walker, 1984). Because complexity goes hand in hand with interorganizational

complementarity, partnering firms need to establish stable cooperation patterns in order to

coordinate the interorganizational “alignment of action” (Gulati et al., 2005): the coordination of

highly interdependent tasks (Gulati & Singh, 1998). Grunwald and Kieser (2007) point to the role

of boundary spanners, e.g. dedicated alliance managers that work on both sides of a relationship.

Cooperation patterns may not only involve more or less standardized procedures to coordinate the

joint production or sales process, the joint project management or the ongoing training and

evaluation of existing partners, but also the selection procedures of appropriate new partners and the

re-selection of old partners (Sarker, Sahaym, & Bjorn-Andersen, 2012; Schreiner, Kale, & Corsten,

2009; see Sydow (2005) in general on network management practices).

Threats and rigidities

Even though the idea of managed complementarity may be attractive, there are also indicators for

threats and rigidities of interorganizational cospecialization processes. To analyze these possible

negative outcomes it is important to highlight the temporal dimension of interorganizational

cospecialization that has not received much attention in empirical studies so far, but that is of

particular interest for our study of interorganizational path dependence. “[C]omplementary assets …

shape going forward enterprise strategy, sometimes positively (in terms of returns to innovation)

and sometimes negatively” (Teece, 2006: 1135). It is important to ask for self-reinforcing

complementarity effects in the processes of cospecialization, because “[p]ure complementarities

without self-reinforcing processes characterize a stable situation of fitting resources, but not a

process that eventually leads to a lock-in” (Sydow et al., 2009: 697). We assume these rigidities on

9

two different levels:

(1) Rigidity in cooperation patterns: Cooperation patterns along the value chain can be seen as

common agreements necessary to reassemble these different but complementary parts. Grunwald

und Kieser (2007: 374) compare this with the coordination of concert musicians who do not need to

know everything about their colleagues’ instruments: “For example, knowledge about musical

notation and the conductor’s standardized movements reduce the communication required between

the members of an orchestra (i.e., specialists in different instruments) when rehearsing a concerto”.

If assets were similar and firms would possess the same step of the value chain, there would be no

need to collaborate – at least from a cospecialization point of view. Thus, the main goal of partners

is not always to acquire each other’s knowledge bases, but to combine the different but

cospecialized assets to jointly commercialize a more innovative product (Grant & Baden-Fuller,

2004; Grunwald & Kieser, 2007). Therefore, a certain degree of stability is necessary. Templates of

cooperation between different firms along the value chain may emerge in unique historical contexts,

but positive feedback of cospecialization processes may stabilize these templates. What is more,

these cooperation patterns may rigidify and became as source of path dependence. The majority

view in alliance research suggests that in interorganizational relationships coordination patterns are

easier to adapt (Gulati et al., 2005). There are only a few studies that found evidence for inertia in

some specific kinds of interorganizational routines, e.g. the selection procedures of firms in

strategic alliances (Li & Rowley, 2002). We aim to go a step beyond these findings and argue that

interorganizational cooperation patterns can become path-dependent in a way that they cannot be

adapted to environmental change or to the diverse demands of diverse user groups.

(2) Rigidity in asset specialization: The division of labour between firms also causes local learning

processes. The different partnering firms with their bounded rationality (Simon, 1979) and limited

resources engage in specialist learning and mainly focus on the development of their own assets

(Grant & Baden-Fuller, 2004; Grunwald & Kieser, 2007). Therefore, firms’ assets might become

increasingly different as they are becoming increasingly cospecialized. Especially when we think of

assets and capabilities that are characterized by ongoing change, specialization or life cycles

(Hansen, McDonald, & Mitchell, 2013; Helfat & Peteraf, 2003). These local learning processes may

foster the threat of specialized “silos” that can inhibit systemic innovation (Jacobides & Winter,

2005). Weigelt (2009), for instance, finds that outsourcing can negatively affect the performance of

a firm, because firms are losing context-specific knowledge. Furthermore, extensive vertical

disintegration may also inhibit systemic innovation (Chesbrough & Teece, 1996). A history of

interorganizational complementarity effects may have the consequence that assets are unequally

specialized across organizational boundaries (cospecialization) and cannot be re-coordinated

10

quickly enough to match new environmental demands.

Path dependence in complementarity-based alliances

Based on the existing theoretical insights on cospecialization as well as the model of organizational

path dependence by Sydow et al. (2009), we conceptualize the following four stages of path

dependence in complementarity-based alliances.

Phase I: From the initial preformation phase with real alternatives, a specific complementarity-

based collaboration mode is chosen by a focal firm and its partner firms. Sometimes activities that

were performed jointly within one organization, become separated activities that are rendered by

different vertically related firms along the value chain (Jacobides & Winter, 2005). This separation

of assets and the outsourcing to a partner firm or a partner network is an important prerequisite for

the efficacy of interorganizational complementarity effects: “On a more general level,

complementarities mean synergy resulting from the interaction of two or more separate but

interrelated resources, rules, or practices” (Sydow et al., 2009: 699, emphasis added).

In phase I, interorganizational cooperation patterns are not yet stabilized, but when firms start to

increasingly focus their learning processes on different cospecialized assets, this may lead to

complementarity effects that gain momentum and a critical juncture (Collier & Collier, 1991) leads

to the next phase.

Figure 3: A model of interorganizational path dependence

Phase II: In the path formation phase, the self-reinforcing complementarity effects render increasing

returns, relational advantages (e.g. through installed base advantage) and the interorganizational

11

cooperation pattern stabilize. These common templates allow both sides (focal firm and partner

firms) to reduce interpartner learning and to further specialize in the development of their

increasingly different but increasingly complementary assets along the value chain (see Figure 3).

Phase III: A lock-in occurs, when the focal firm's attempts to form and stabilize alternative

interorganizational cooperation patterns fail. This is because historically emerged cospecialized

assets are locked into “silos” and cannot be replaced or re-coordinated quickly enough across

organizational boundaries. Both sides fall back on old ways of coordinating that are (at least

potentially) inefficient for the focal organization.

Phase IV: The theory of organizational path dependence states that organizational lock-ins do not

last forever, but can be broken on a discursive, behavioral, systemic or resource-based level

(Schreyögg, Sydow, & Koch, 2003). Similar to the emergence of path dependence, the breaking of a

path occurs as an exceptional, emergent and not fully predictable process (Dobusch & Kapeller,

2013; Dobusch, 2008; Plowman et al., 2007). The same can be assumed for the breaking of

interorganizational path dependencies, but little is known about this issue. Drawing from our case

study we shed light into this phase of path dependence.

Research setting and design

We investigate the research problem with a case study of SAP’s initiative to enter the lower SME

market with a new product and partner strategy. The partnerships between SAP and its service

partner firms are a perfect example of cooperation based on complementary (cospecialized) assets

(Teece, 1992). This mode of collaboration in the ERP industry has been described as an example of

value co-creation (Sarker et al., 2012). It demands coordinating mechanisms like selection,

evaluation and training procedures and dedicated alliance managers on both sides (Schreiner et al.,

2009). Furthermore, the cooperation pattern between SAP and partners becomes deeply inscribed

into sociomaterial entanglements, such as a common programming language and a specific software

parameter logic. These software parameters can be thought of as programmed switches that are used

by consultants to implement and individually configure a copy of a standardized SAP ERP package.

A parameter logic allows SAP on the one side to continuously increase the functionality of the

software package by adding further parameters. The service partners on the other side can use these

parameters to customize the software package to specific customer needs (Lehrer & Behnam,

2009). SAP’s very comprehensive and complex ERP software (formally known as R/3) comprises

approx. 10,000 parameters and several modules.

Through this offer and in cooperation with partner firms (e.g. for sales and software

12

implementation), SAP has become the undisputed worldwide market leader among large enterprise

customers. For the SME market, however, this product strategy as well as the respective partner

model misfit and cause strategic inefficiencies. This is why SAP has been trying to heavily

restructure its product and partner model for the SME market since 2007. A new web-based

software called Business ByDesign has been developed. The new software should be purchased and

installed by customer-self-service via the Internet (software-as-a-service). As shown in Figure 4,

only 300 parameters should be enough to allow the customer to simply set up this software (Faisst,

2011). Nevertheless, to fulfill the different needs of individual SME customers, it was planned later

to engage partner firms to develop and distribute add-on-software on SAP’s commercial online

platform (platform-as-a-service). Until today, this initiative has fallen way short of expectations,

because in contrast to the plan of simple customer-self-service the software is delivered in a

complex large enterprise fashion, still demanding extensive consulting and customizing services by

partner firms.

Figure 4: Failed restructuring of SAP’s product and partner model and value chain

As we are interested in dynamics of interorganizational processes that are difficult to demarcate, we

choose a case study design (Yin, 2009). More precisely, we investigate our research question with a

holistic case study focusing on two critical episodes. This temporal bracketing of process data “can

be especially useful if there is some likelihood that feedback mechanisms, mutual shaping, or

multidirectional causality will be incorporated into the theorization” (Langley, 1999: 703).

The first part is a historical analysis of the rise of the ERP software vendor SAP’s partnering and

innovation strategy in the last four decades (R/3 episode). We analyze how SAP became the world's

13

biggest vendor of enterprise software with a network of about 10,000 partnerships (RAAD, 2012).

Most relevant for this process are the so-called service providers, because their activity is very

consulting-intensive (SAP, 2012). Leading service providers are, for instance, Accenture,

Capgemini, PriceWaterhouseCoopers or T-Systems.

The second part will investigate SAP’s recent initiative to enter the SME market with a new product

and partner model from 2007 until 2013. Our analysis focuses not only on SAP but also on partner

firms like IMPLEMENT, which is SAP’s most innovative Business ByDesign partner network

(Business ByDesign episode / short: BBD episode). This research design allows us to relate current

rigidities to historical roots.

Methods of data collection, coding and analysis

For our analysis we use archival data, data retrieved from documents, participant observation, as

well as interviews. We conducted 25 interviews with managers of SAP partner companies, 6

interviews with SAP customers, and 8 interviews with managers at SAP. The gathered data was

organized in a qualitative case study database to improve reliability (Yin, 2009). We gathered

written data (transcripts, protocols, Excel sheets, and presentations), audio data of workshops,

internal meetings, and telephone calls as well as research notes from participant observation

between 2011 and 2013. What is more, we triangulated our data with findings of previous studies

and press reports on SAP and its partners.

In the first step, we reconstructed the history of both the R/3 episode and the BBD episode as a

written thick description (Geertz, 1973). Then we analyzed why SAP has not been able to

restructure its product and partner model for the SB/SME market and we applied the concept of

path dependence to the case, e.g. relating present developments to causes in the past. More

precisely, we searched the data for concrete competencies and practices that were meant to be

changed, but that were resistant to change. We found that the specific way of coordinating tasks

between SAP and partners became a stable mechanism in the R/3 episode, enabling complementary

value creation and cospecialization. We used the software MAXQDA for coding (data from German

sources was translated into English) and started the coding process from two directions. On the one

hand we searched the data for concrete competencies and practices in the interorganizational setting

that were meant to be modified, but that were resistant to change (empirical codes). We also used

empirical codes to analyze barriers to change, critical events and other relevant contextual

parameters. On the other hand we simultaneously looked for empirical support for theoretically

derived indicators from the path dependence literature (theoretical indicators taken from Burger &

14

Sydow, 2012; Koch, 2008; Sydow, Windeler, Müller-Seitz, & Lange, 2012). Through this

interdependent coding process, an interpretative coding scheme was iteratively developed that fitted

to the general theoretical assumptions of the theory of path dependence as well as to the empirical

setting of interorganizational cospecialization and cooperation (see Figure 5). This coding strategy

is neither a grounded strategy approach (Glaser & Strauss, 1967), nor a deductive confirmation of

theoretical hypothesis, but can best be described as “pattern matching” approach (Yin, 2009) for the

purpose of theory development (extending the theory of organizational path dependence to

interorganizational settings).

Figure 5: Development of an interpretative coding scheme

We use Koch’s (2011) conceptual distinction between “on-path” and “off-path” to stress the fact

that managers might have clear concepts of how to unlock the path through change initiatives

(getting from on-path to off-path). However, the dynamics of self-reinforcing mechanisms may lead

to a relapse to the old cospecialization and cooperation patterns (which is path-relapse). Table 1

provides illustrative evidence for each of the interpretive codings:

15

Phase Interpretative

code

Example

On-path (the path-dependent pattern)

Competencies SAP

“Realtime, online and standardization. These were the unshakable elements of SAP. Because of the massive growth, we couldn’t do everything alone. […] It would have been too risky to hire so many consultants” (Interview, SAP, manager, 71).

Cooperation pattern

“Then we have trained our partners and their employees. Of course, when they do mistakes within an implementation project, people don't say the partner is bad, but they say SAP is bad. This is why we have put extreme value on the quality of the partners' employees, that they are well trained” (Interview, SAP, manager, 90).

Competencies SAP partners

“Today, the auditing firms are good SAP experts. Otherwise they couldn't audit” (Interview, SAP, manager, 115).

Strategic target “The R/3-world was characterized by a focus on large enterprises, corporations [...], companies with more than 1.000 employees” (Interview, SAP partner, manager, 46).

Off-path (the change initiative)

Competencies SAP

“The objective of this new approach is to significantly reduce the effort and length of implementation” (Faisst, 2011: 29).

Cooperation pattern

“In the beginning they had that way of thinking that you go to the portal, identify yourself as belonging, for example, to the retail industry and, immediately, your system is set up” (Interview, industry expert, 99).

Competencies SAP partners

“In the beginning we thought that we need partners with this Cloud-DNA, someone who can somehow sell a cloud” (Interview, SAP, manager, 79).

Strategic target “The strategic thrust focused the SB/SME market” (Interview, industry expert, 90).

Path-relapse (relapse to old patterns despite change initiative)

Competencies SAP

“SAP is using a technology called NetWeaver. The ABAP-Stack belongs to that technology and NetWeaver is very capable but also needs a lot of resources. That means Business ByDesign was extremely slow in the beginning, because it inherited the whole complex stack” (Interview, industry expert, 18).

Cooperation pattern

“And then they had to realize 'OK, we also need add-ons' and they said 'Maybe we need partners after all'” (Interview, industry expert, 99).

Competencies SAP partners

“But the primary assumption that Cloud needs a totally different type of partner is not true from my point of view, because the complexity to sell the solution remains the same after all” (Interview, SAP, manager, 5).

Strategic target “Silently SAP has [...] raised the target user numbers with that release change and focused this segment with direct sales to large enterprises” (IT-Reseller.de, 2012).

Table 1: Interpretative codings

16

Findings

The success spiral of SAP’s product and partnering strategy

SAP was founded in 1972 by five former IBM employees. They were among the first companies

that developed standardized instead of individual enterprise software. As requirements towards

enterprise software differ between organizations, it was not a trivial endeavor to develop software

that is highly standardized and at the same time suitable to idiosyncratic requirements within

different organizations. SAP’s software solved this trade-off between standardization and

adaptability by using a design principle called programmability (Lehrer & Behnam, 2009). The

shipped SAP software is so comprehensive and abstract in functionality that it can and has to be

customized by consultants setting the parameters to specific requirements.

“The complexity of the undertaking exceeds the capability of company IT departments, thus requiring

SAP or (more often) an external IT service or consulting company to supervise implementation of the

ERP system” (Lehrer, 2006: 197).

This technological separation of SAP software and services started back in the 1970s and allowed

SAP to transfer the service job to partner firms (see Figure 6).

Figure 6: Separation of the value chain and vertical desintegration

This led to the rise of the SAP service industry in countries like the US, France, UK or Germany

(Lehrer, 2006).

“The concept of customizing in R/3 was revolutionary, also in R/2. That is the reason why SAP exists,

because SAP was the first one who developed customizable software” (Interview, SAP, Manager, 14).

17

Even though there were many companies, especially in Germany, trying to develop standardized

enterprise software packages, SAP was the only company that consequently transferred the service

job to partner firms.

“The reason why SAP’s strategy was different from competitors’ strategies [...] was the consequent

realization of standardization in the R/2-development. The hereby achieved abstraction demanded a more

or less extensive consulting during implementation in the concrete user firm and firm or industry specific

supplements. This task could not be achieved by SAP alone and therefore it was reasonable to look for

consultancy firms as partners – especially because these firms often had the necessary know-how that

SAP did not have or could not build up” (Leimbach, 2007: 46).

Standardized SAP-software and the separated services are complementary in the sense that the

better they fit to each other the better the result for certain customer segments. By this self-

reinforcing mechanism the success spiral in the large customer segment gained its momentum. The

multi-level value chain did not only increase the distribution possibilities (during expansion partners

provide the necessary personnel very quickly). This complementarity effect of cospecialization also

increases the application possibilities of the product. SAP and SAP partners learn with each new

customer project: New knowledge is transferred back into the program code of the standardized

software and each project also enhances the knowledge of the SAP partners. This qualitative

improvement in turn reinforces the attractiveness for certain customer segments.

Interorganizational cooperation patterns

SAP and its partners could harvest the benefits of interorganizational cospecialization because they

found suitable practices of interorganizational coordination. After dividing enterprise software in

two distinct elements (development of standardized software packages and the customization of this

standardized software by partners), these two elements needed to be coordinated. Therefore, SAP

established a partner management that was responsible of selecting, evaluating and training the

service partners. The partner firms on the other side also mirrored this cooperation pattern by

engaging dedicated SAP partner managers, developers and consultants.

The cooperation pattern reinforces the creation of complementary assets (e.g. the specific role

ascription: SAP as software developer, partner firms as service providers). Therefore, when SAP

selects or evaluates partners, an overall objective is the reinforcement of the service capability of

partners:

“A partner who wants to provide services in the field of SAP ERP has to attend many certification courses

and all of his employees are certified. One of our metrics to evaluate the partner ecosystem is the number

of certified SAP consultants at the partner firm, because this means they are particularly good,

particularly well qualified” (Interview, SAP, Manager, 14).

18

During the years the requirements towards partners grew and their role as service partners was

reinforced:

“Because of their permanently changing software releases, SAP’s requirements towards their partners are

very high. You need to permanently train your workforce. We are gold partner. This brings the advantage

that we have a close contact to SAP [...]. They invite us to test new developments and we are also asked to

bring in new ideas” (Interview, traditional SAP partner, Manager, 44).

On the technical level, an increasing amount of software parameters constantly allowed a more

advanced coordination between SAP and service partners. Beside these technical means, complex

partner selection, evaluation and training procedures enabled a well-coordinated strategy of

interorganizational asset complementarity and together with its partner firms SAP became the

largest enterprise software vendor in the world. The revenue grew from almost 92 Mio. € in 1988,

when SAP went public, to 11,575 Mio. € in 2008. In the same time the number of employees grew

from 940 to 51,400 (Leimbach, 2007; SAP, 2008a). Simultaneously, the number of SAP consultants

grew up to 160.000 in 2008 (PAC, 2009). Most of these consultants work for SAP service partners

or at user firms. In the following, the world’s largest SAP service partners are listed – each with the

year of the first collaboration with SAP in brackets: IBM (1972), Accenture (1975), CSC (1979),

BearingPoint (1982), SIS (1982), Capgemini (1985), Atos Origin (1989), Ciber (1989), HP (1990),

T-Systems (1990), Reply (1992), Mahindra (1995), MHP (1996), Infosys (1998), itelligence (1998),

msg systems (1998). Meanwhile, “[t]hese system integrators (SI) have grown substantially with

SAP and SAP has grown with the SIs, too” (Meyer, 2008: 11).

The flip-side of SAP’s product and partnering strategy

Quite early, the success of SAP’s product and partnering strategy revealed a flip-side. In the large

enterprise segment complexity of the R/3-ERP-software was more of a competitive advantage than

a disadvantage, because it meant a wider range of applicability. Complexity was also the driver of

the business model of partner firms, because they gained their revenue with customizing and

consulting services. But when SAP’s growth potential in the large customer segment diminished

due to market saturation, the tremendous product complexity and the costs for service partners

hindered the exploitation of upcoming market potential in lower market segments.

“SAP's use of programmability to resolve the standardization-adaptation tradeoff came at the cost of

increasing product complexity and hence difficulty of implementation by users …” (Lehrer & Behnam,

2009: 286).

Various failed SME initiatives underpin the assumption that SAP’s partner and product strategy did

not fit to the SME market (Eisenhardt, 2002). Until today all SME initiatives had to struggle with

19

the legacy of SAP’s chosen path. First attempts to enter the SME market can be traced back to the

late 1980ies, when SAP built stakes in different service firms that had an expertise for the SME

market. But the acquisition model turned out to be a failure and this is why SAP later tried to enter

the SME market with a partner model (Computerwoche, 1995). In 1995, the Systemhausinitiative

was launched. Systemhaus is a German term for (often small) service providers in the IT industry.

With their industry expertise these new partners tried to reduce the complexity of SAP’s complex

software package, programming SME industry solutions. Partial success could be achieved by this

strategy among larger SME customers, but the structural problem within the lower SB/SME market

could not be resolved (Computerwoche, 1995, 1996, 1997). This also applies to the product

relaunch during the new economy boom between 1999 and 2001 (Computerwoche, 1999, 2001).

The most important change that could be achieved within the SME initiatives was that partner firms

were engaged to pre-customize the R/3-software for certain midsize firms with 1,000 and more

employees in sub-industries. By this, some market share in the upper midsize market could be

achieved. However, despite an extreme effort, SAP is a long way from reaching a share in the SME

market comparable to that in the segment of large customers. In fact, the interorganizational

cooperation pattern has not been changed substantially for the upper midsize market. The technical

platform remains R/3 and the parameter logic is used to coordinate SAP’s software development

and the customizing by service partners. The role of service partners in the upper midsize market is

also mainly the role of service partners.

Figure 7: German market for business software (SAP, 2008b: 5)

In the SME market the demand for simplicity and efficiency collides with the consulting-intensive

SAP offerings. This is the reason why even after more than 20 years of SME initiatives SAP is far

from reaching as high market shares in the SME market as in the large customer market. Figure 7

illustrates the German market for enterprise software in 2008, broken down by customer size of

corporate customers. SAP customers are shown in gray, the untapped markets in white.

20

SAP’s attempt to diversify product and partner strategy

The constant problem in the SME market became even more relevant in the 2000s, when the cloud

computing trend emerged. The idea of cloud computing is that IT resources are not installed on the

customer's premise any more but accessed via the Internet. Software that is accessed in a cloud

computing fashion is called software-as-a-service, because the customer does not pay for licenses,

implementation and maintenance separately, but with a monthly subscription fee instead (Yang &

Tate, 2012). The launch of SAP Business ByDesign addressed this radical new approach. In the

long run, this software was intended to replace the (R/3-based) ERP platform that has been

continuously enhanced, but not fundamentally changed since it was introduced in 1993 as R/3.

Because it was supposed to become the “new R/3”, Business ByDesign has important strategic

significance for SAP. Against this backdrop, Henning Kagermann, SAP’s CEO in 2007, commented

on the corporate strategy as follows:

“I have been at SAP for 25 years, and this is the most important announcement I have made in my career

here […]. We designed this product to launch a new business model. It's a new way to design, develop

and implement business software” (Kawamoto, 2007).

Bounded change: Reintegration through automatization

SAP managers were aware of the fact that product complexity and the high costs for SAP

consultants had impeded the success in the lower SME market. Thus, the plan was to automatize

services and sales processes by integrating these steps into the new Cloud Computing software.

“We said, we build the implementation method into the product” (Interview, SAP manager, 4).

“When I came to SAP, they said 'It will be implemented online by customer self-service and paid by credit

card'” (Participant observation, SAP, 1).

Looking at the value chain, this plan meant nothing less than a significant reintegration of activities

that had been vertically disintegrated within former cospecialization processes. The plan was to

develop “consulting-free” software: easy ERP solutions that could be purchased and set up by SME

customers via Internet-self-service.

Asynchronous change: New interorganizational cooperation patterns

However, after the first product announcement in 2007, it turned out that the new sales and service

approach did not meet the SME customers’ needs. The new product is more than a low-weight line

of business solution; it is a fully integrated ERP suite with different modules from financials and

human resources to compliance and project management. This is why SAP experts are necessary

after all – for implementation and for the sales process. Therefore, SAP started to develop a new

21

partner program between 2007 and 2010. Inside of SAP, discussions of how to change the

coordinating with partners took place.

“There were almost ideological discussions whether to commercialize this new product with or without

partners. Hasso Plattner, our chairman [and SAP co-founder] was very engaged in that project during

that time and he wanted to sell this product without partners, because a web-based product is predestined

for a web-based direct distribution” (Interview, SAP, Manager, 14).

In this regard, there were attempts to overcome the underlying semantics of the old coordinating

mechanism:

“A part of this project was to question SAP's existing concepts and the terminology was adapted,

especially where there were negative connotations. 'SAP implementation', for example, has the

connotation 'Lasts forever and is unbelievably expensive'. The same is the case for the word 'customizing'

[...]. This is why it was prohibited for some time to talk internally about implementation or customizing.

When you had these words on your slides you were kicked out of the presentation. This is why we spoke

only about 'go-live'”(Interview, SAP, Manager, 14).

As the new software should have been implemented by customer self-service as far as possible, the

question aroused whether the established partner model could work for the new strategy at all. After

internal discussions, SAP assigned a new role to their partner firms. They should primarily develop

and distribute “add-ons” for SME customers on an online “platform-as-a-service” (Faisst, 2011).

This online development was meant to become the new way of adapting standard software-as-a-

service to customers' individual needs. The online platform was also planned to become a new

practice of technical coordinating, because it offers an online development environment for SAP

partners. The partner selection, evaluation and training procedures were also planned to become

simple online processes with a minimum of bureaucracy.

However, the promising SAP Business ByDesign initiative has failed until now. It was planned to

gain 10,000 SME customers and a turnover of 1,000 Mio. € in 2010, but in 2011, only 1,000

customers could be convinced and a total turnover of 18 million € was achieved (stern.de, 2012).

Although rapidly gaining a critical mass of partners and customers would have been important to

forge a new form of asset complementarity and to profit from a new technological path, the

opposite has happened.

Relapse to established interorganizational cooperation pattern

Companies like salesforce.com are very successful in distributing simple enterprise software-as-a-

service to large and small customers. They also found innovative practices of coordinating with

partners in a platform-as-a-service model. In other words, there were success stories in the

enterprise software market. By contrast, SAP’s attempt to develop simple software-as-a-service

22

failed. Coming from the large enterprise world and used to the coordinated collaboration with

highly specialized consulting firms, the new product Business ByDesign was developed in such a

comprehensive and complex way that it still needs a lot of consulting by service partners.

Therefore, the original plan to distribute Business ByDesign without customization has failed and

the new practices of coordinating have not come into life. This has also thwarted the plan to sell

software-as-a-service without partners via the Internet. Similar to the large enterprise segment,

partners have been engaged as service providers who adapt the product complexity to individual

customers’ needs with consulting and customization. Additionally, because SAP could not reach

small and medium enterprises via the Internet, they also had to assign the role of offline resellers to

partner firms that engage in complex sales cycles. Contrary to the plan, all these difficulties have

led to a relapse to old complex partner selection, evaluation, training procedures, because well

trained partner firms were needed for these complex challenges.

Our interview partners frequently compared SAP’s coordinating practices with the partner

management of “web-born” companies like salesforce.com or Google. Unlike SAP, they have been

very successful in efficiently coordinating with partners in the online business, whereas SAP’s

selection and evaluation procedures are still imprinted by the old bureaucratic coordinating

mechanisms:

“There are still many formal barriers at SAP. They have the wrong contracts, wrong contractual

constructs with inappropriate conditions for a partnership with COMPLEMENT” (Interview, New SAP

partner COMPLEMENT, Manager, 1).

Furthermore, SAP has not been able to find new “web-born” partner firms for that new business.

Even the most innovative SAP partner network dedicated to Business ByDesign consists to a large

extent of old SAP consultants. Originally, it was planned to acquire consultants who are not

“spoiled” by the old R/3 consulting business.

“There are ten or twenty thousand consulting partners in the old saturated SAP business in Germany [...].

Sometimes it is better to find partners that are web-born and who are not from the world of SAP large

enterprise projects” (Interview, New SAP partner IMPLEMENT, Manager, 108).

Additionally, in contrast to SAP’s plan, we observed that partner firms still receive their old role

instructions in the daily communications with SAP. Even though it was planned to engage partner

firms in new roles as online add-on-developers, in the daily practice they are still encouraged to sell

services on premise and to adjust their business model to consulting.

From bounded and asynchronous change to path-relapse

Within the Business ByDesign initiative SAP attempted to unlock the path-dependent

23

interorganizational cooperation pattern that had been stabilized by

cospecialization processes with service partners in the large enterprise market. The plan was to

integrate the value creation activities (e.g. services) into a new Cloud software through

automatization (bounded change, see Figure 8).

Figure 8: Change initiatives and path-relapse

However, the complexity of ERP implementation and sales still required the work of SAP experts

and between 2007 and 2010 a new partner program was developed by SAP (asynchronous change).

In particular, partners should receive new roles that should fit to the new Cloud paradigm. Yet, in

practice the new concepts of the new partner program were not adopted by the staff at SAP partner

firms and at SAP and were quickly replaced by existing concepts that have proofed their worth in

the established SAP ecosystem.

Discussion

In this study we show why SAP has been unable to unlock interorganizational path dependence. The

interorganizational asset cospecialization processes that set in motion a success spiral in the large

enterprise market stabilized an interorganizational cooperation pattern that cannot be disrupted for

new environments. Thus, SAP’s interorganizational cooperation pattern has become path-

dependent. According to Sydow et al. (2009) path-dependent processes are characterized by three

different stages, a phase of contingency, a phase of path formation and a phase of lock-in.

Indicators for all three different phases can be found in the case at hand. At first, there were real

24

alternatives. The founders of SAP could have gone the way of their competitors and distribute and

implement their software without service partners. Secondly, after the partnering path had been

taken, the increasing interorganizational complementarity effects led to further stabilization of the

interorganizational cooperation pattern. In the case at hand, the cooperation pattern increasingly

relied on a specific interorganizational role ascription that SAP cospecializes in software and

partner firms cospecialize in the complementary services. Thirdly, the interorganizational

cooperation pattern is locked-in in the sense that SAP cannot engage old or new partners in new

roles, e.g. as online add-on developers. The interorganizational alignment of action in the new cloud

business is different from the established ways of coordinating. SAP managers were aware of this

fact and developed new practices of coordinating that should differ fundamentally from the old

ways of coordinating. Over the years, there has always been change within SAP’s partering strategy.

Nevertheless, the concept of path dependence (Sydow et al., 2009) can explain why – despite

incremental change – the overarching “architecture” of coordinating could not be substantially

modified. The increasing functionality and complexity of SAP’s software has gone hand in hand

with an increasing cospecialization (Teece, 1992) of SAP partners in software consulting and

customizing – indicating a self-reinforcing complementarity effect. The process has increased the

attractiveness of SAP’s offerings for large corporate clients with a high demand for

consulting/customizing services. Yet at the same time, the high expenses that have come along with

these services have decreased the attractiveness for small businesses. After several decades of

successful interorganizational cospecialization, assets have become increasingly diverse as they

have become increasingly cospecialized. This unequal distribution of capabilities across

organizational “silos” (Chesbrough & Teece, 1996; Jacobides & Winter, 2005) makes it ever more

challenging to change the whole value chain at once, heavily reducing the adaptability of

interorganizational cooperation patterns. Without the cooperation with partner firms SAP did not

learn quickly enough to reintegrate formerly disintegrated assets for the development of simple,

end-user friendly software within the Business ByDesign initiative. So they fell back on the old

cooperation patterns, delegating the “complexity reduction job” to partner firms like it had been

reasonably learned in the large enterprise segment.

An important finding of our study is that long-term effects of cospecialization processes between

firms do not only stabilize the interorganizational cooperation pattern and the asset base of the

collaborating firms. Interorganizational cospecialization processes also have long-term effects

within the participating organizations because the dominance of specific cospecialized assets shape

the (bounded) rationality of a firm. Therefore, to fully understand why SAP was unable to unlock

25

their path-dependent interorganizational strategy, we have to extent existing concepts that are

known in the path dependence literature. Within our case study we could identify two traps for path

breaking initiatives that are relevant especially in interorganizational settings. Per definition,

interorganizational relationships consist of different autonomous levels and interorganizational path

dependence that is driven by positive feedback of asset complementarity leads to an increasing

cospecialization (thereby these assets also become increasingly different). Therefore, it is not

surprising that managers focus on their own assets first within change initiatives and it is quite

likely that these path breaking initiatives start with bounded change in only one organization or in

organizational subunits of only some organizations. However, this might be not enough if path-

breaking requires radical and encompassing change. Therefore, we name this problem the (1) trap

of bounded change. In the case of SAP Business ByDesign we could observe an autonomous path

breaking initiative in the beginning. In the first years SAP focused only on their own “silo”, trying

to develop a completely new product in a new subunit without dealing with the partner side. The

other side of the path dependence – the interorganizational dimension – was addressed later.

Figure 9: Failed path-breaking in interorganizational settings

This reveals another problem of path breaking initiatives in interorganizational settings, the (2) trap

of asynchronous change. Asynchronous change is a problem since the cooperating partners cannot

appropriately adjust their highly interdependent business processes (see Figure 9). Meanwhile, if

partners do not coordinate their change patterns early enough, it is likely that the positive feedback

of the old path will quickly neutralize path-breaking dynamics.

Both traps, the trap of bounded change and the trap of asynchronous change, may lead to a relapse

to the old path, especially when the attracting forces of the old path are still very strong or even

growing (see Figure 10).

If the breaking of an existing path dependence is intended in the long run, there are many

indications that the change patterns would have to be comprehensive and synchronous. It is

subunit 1 O O O O O O O O O O O O O O O O O O O O

subunit 2 O O O O O O O O O O O O O O O O O O O O

subunit 3 N N N N N N N N N N N N N O O O O O

subunit 3 N N N N N N N O O O O O

subunit 2 O O O O O O O O O O O O O O O O O O O O

subunit 1 O O O O O O O O O O O O O O O O O O O O

t

O = old path / N = new path

organization A

Bounded change

Relapse to old path

organization(s)

B, C …

Bounded change

Asynchronous change

26

questionable if this kind of de-locking, which demands radical (inter-)organizational change

(Greenwood & Hinings, 1996), as well as radical (Henderson & Clark, 1990) or disruptive

(Christensen & Raynor, 2003) product innovations can be achieved within a path dependent value

chain that is driven by the complementarity effects of cospecialization: “That is, excessive and

sustained specialization may create ‘silos’ that inhibit systemic business improvement“ (Jacobides

& Winter, 2005: 404).

Figure 10: Path-breaking versus path-relapse

SAP still has its thriving large enterprise business with established complementarity effects that

serve as a cash cow and will so for several years. In a new market with a new product, however, the

critical mass of new customers could not be reached. It seems that the winner that takes all in one

market can sometimes not win much in very different markets. This has implications for

diversification strategies of firms in digital markets and other domains with high network effects

where complementarity-based alliances are common. As Stieglitz and Heine (2007: 13) put it,

diversified firms “... desire to fully exploit existing complementary assets across different markets

with the need to establish specialized activity systems for each one”. Thus, the establishment of new

cooperation patterns can be very challenging.

Conclusion

Companies utilize interorganizational cospecialization in order to gain relational advantages. The

coordination of such relationships that reach beyond organizational boundaries can be challenging.

Firms are sometimes unable to adapt these interorganizational arrangements (e.g. for the access to

new markets). We have analyzed this problem by focusing on a qualitative case study of SAP’s

effort to enter the market of small and medium enterprises which had been a more or less untouched

in recent years and which is known to be a difficult terrain. SAP decided to enter this market with a

new product and partner strategy in recent years. We have precisely reconstructed the period

between 2007 until 2011 through interviews and archival data. Between 2011 and 2013 we have

27

accompanied the process using participant observation. We have compared the findings with a

historical analysis of the rise of SAP’s product and partner strategy in the last four decades. This

allowed us to relate current rigidities to their historical origins. More precisely, we have presented

evidence for the relapse to the old, alleged success-path and based on the empirical evidence we

have discovered bounded change and asynchronous change to be the critical traps. Thus,

cospecialization and cooperation patterns which emerged historically became path-dependent and

therefore path-breaking fails.

To avoid the failing of path breaking initiatives, managers should reflect on the specific aim of their

initiatives. Different approaches are necessary, depending on whether path-breaking or path-

coexistence is the aim of the change initiative. It is possible to establish a new path that coexists (for

example a new product line) with an existing path, when the attracting forces of both paths do not

interfere with each other. SAP, for example, acquired the Business Intelligence Software called

Business Objects in 2007. This product line does not interfere with the core business of SAP, but is

a logical supplement. The initiative of Business ByDesign, however, has strong interferences with

the core business of SAP. But it was not clear, if the new product and partnering strategy should

substitute the old product and partnering strategy in the long run (path-breaking), or if the Business

ByDesign path should be run parallel to the R/3 path (path-coexistence). In the course of time, a

partial path-relapse emerged, because of the traps of bounded and asynchronous change (see Figure

10 once again).

Ultimately, our study has contributed to the research on path dependence adding new insights from

interorganizational alliances. This comprises not only a theoretical integration of path dependence

theory with the concepts of cospecialization and complementarity effects but also empirical

evidence from our explorative case study. Thereby, we extended the existing knowledge about path-

breaking failures. More precisely, we have identified two critical conditions for path-breaking

failures. Despite this contribution, our study has several limitations that should be mentioned. Just

as in all (single) case studies the issue external validity can be raised. In this study, the historical

development of a very specific organization and its relationships to external partners is analyzed.

These interorganizational arrangements historically evolved in many years and led to a certain

structural embededdness of the focal firm SAP and its partners. The exploratory approach and the

qualitative data offer starting points for criticism towards the question of internal validity and

reliability. Even though these limitations can never be resolved completely, the triangulation of

several sources and references to internal and external data sources as well as the coding and

interpretation of the data by two researchers with different expertise should by and large ensure the

28

quality of the methods used. Additional research may contribute to substantiate our findings. For

example, cases with successful path-breaking initiatives in the same/similar industries could be

valuable as a contrast to our case study. Most relevant for this is Figure 9, which demonstrates the

identified traps of path-breaking and could be further examined and validated. Also, it might be a

challenge to assess this phenomenon quantitatively which, however, will be a difficult undertaking

due to the dynamic and interdependent processes that are involved over time.

Literature

Arthur, W. B. (1989). Competing Technologies, Increasing Returns, and Lock-in by Historical

Events. Economic Journal, 99(394), 116–131.

Arthur, W. B. (1994). Increasing Returns and Path Dependence in the Economy. Ann Arbor:

University of Michigan Press.

Burger, M., & Sydow, J. (2012). How Inter-Organizational Networks Can Become Path-Dependent:

Bargaining in the Photonics Industry. Schmalenbach Business Review (2014), 66. Retrieved

from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2125188

Chesbrough, H. W., & Teece, D. J. (1996). When Is Virtual Virtous? Organizing for Innovation.

Harvard Business Review, 74(1), 65–73.

Christensen, C. M., & Raynor, M. E. (2003). The Innovator’s Solution: Creating and Sustaining

Successful Growth. Boston, MA: Harvard Business Review Press.

Chung, S., Singh, H., & Lee, K. (2000). Complementarity, Status Similarity and Social Capital as

Drivers of Alliance Formation. Strategic Management Journal, 21(1), 1–22.

Collier, R. B., & Collier, D. (1991). Shaping the Political Arena: Critical Junctures, the Labor

Movement, and Regime Dynamics in Latin America. Princeton, NJ: University Press

Princeton.

Colombo, M. G., Grilli, L., & Piva, E. (2006). In Search of Complementary Assets: The

Determinants of Alliance Formation of High-tech Start-ups. Research Policy, 35(8), 1166–

1199.

Computerwoche. (1995). Partnerkonzept sieht Branchenunterstuetzung vor SAP versucht sich noch

einmal am Mittelstand - computerwoche.de - Archiv 1995 / 6. Retrieved January 19, 2012,

from http://www.computerwoche.de/heftarchiv/1995/6/1112136/

Computerwoche. (1996). Systemhaus-Geschäft soll ausgebaut werden: Im Mittelstand kommt SAP

nur langsam voran - computerwoche.de - Archiv 1996 / 39. Retrieved January 19, 2012,

from http://www.computerwoche.de/heftarchiv/1996/39/1109433/

29

Computerwoche. (1997). “Wir müssen mit einem Schuß ins Schwarze treffen” - computerwoche.de

- Archiv 1997 / 13. Retrieved January 19, 2012, from

http://www.computerwoche.de/heftarchiv/1997/13/1097840/

Computerwoche. (1999). SAP lehnt sich weit aus dem Internet-Fenster - computerwoche.de -

Archiv 1999 / 38. Retrieved January 19, 2012, from

http://www.computerwoche.de/heftarchiv/1999/38/1088830/

Computerwoche. (2001). Bringt Mysap.com den Durchbruch für das Mittelstandsgeschäft?: SAP-

Systemhäuser zwischen den Fronten - computerwoche.de - Archiv 2001 / 11. Retrieved

January 19, 2012, from http://www.computerwoche.de/heftarchiv/2001/11/1064058/

David, P. A. (1985). Clio and the Economics of QWERTY. The American economic review, 75(2),

332–337.

Dobusch, L. (2008). Windows versus Linux: Markt-Organisation-Pfad. Wiesbaden: VS Verlag für

Sozialwissenschaften.

Dobusch, L., & Kapeller, J. (2013). Striking New Paths: Theory and Method in Path Dependence

Research. Schmalenbach Business Review, 65(in print).

Dyer, J. H., & Singh, H. (1998). The Relational View: Cooperative Strategy and Sources of

Interorganizational Competitive Advantage. The Academy of Management Review, 23(4),

660–679.

Eisenhardt, K. M. (2002). Has Strategy Changed? MIT Sloan Management Review, 43(2), 88–91.

Ennen, E., & Richter, A. (2010). The Whole Is More Than the Sum of Its Parts— Or Is It? A

Review of the Empirical Literature on Complementarities in Organizations. Journal of

Management, 36(1), 207–233.

Faisst, W. (2011). Die nächste Generation der Unternehmens-Software am Beispiel von SAP

Business ByDesign. Wirtschaftsinformatik & Management, 2011(4), 24–30.

Gawer, A. (2011). Platforms, Markets and Innovation. Cheltenham (UK): Edward Elgar Publishing.

Geertz, C. (1973). The Interpretation of Cultures. New York: Basic Books.

Gilbert, C. (2005). Unbundling the Structure of Inertia: Resource Versus Routine Rigidity. Academy

of Management Journal, 48(5), 741–763.

Glaser, B. G., & Strauss, A. L. (1967). The Discovery of Grounded Theory: Strategies for

Qualitative Research. London: Wiedenfeld and Nicholson.

Grant, R. M., & Baden-Fuller, C. (2004). A Knowledge Accessing Theory of Strategic Alliances.

Journal of Management Studies, 41(1), 61–84.

30

Greenwood, R., & Hinings, C. r. (1996). Understanding Radical Organizational Change: Bringing

Together the Old and the New Institutionalism. Academy of Management Review, 21(4),

1022–1054.

Grunwald, R., & Kieser, A. (2007). Learning to Reduce Interorganizational Learning: An Analysis

of Architectural Product Innovation in Strategic Alliances. Journal of Product Innovation

Management, 24(4), 369–391.

Gulati, R. (1995). Social Structure and Alliance Formation Patterns: A Longitudinal Analysis.

Administrative Science Quarterly, 40(4), 619–652.

Gulati, R., Lawrence, P. R., & Puranam, P. (2005). Adaptation in Vertical Relationships: Beyond

Incentive Conflict. Strategic Management Journal, 26(5), 415–440.

Gulati, R., Nohria, N., & Zaheer, A. (2000). Strategic Networks. Strategic Management Journal,

21(3), 203–215.

Gulati, R., & Singh, H. (1998). The Architecture of Cooperation: Managing Coordination Costs and

Appropriation Concerns in Strategic Alliances. Administrative Science Quarterly, 43(4),

781–814.

Hagedoorn, J. (1993). Understanding the Rationale of Strategic Technology Partnering:

Interorganizational Modes of Cooperation and Sectoral Differences. Strategic Management

Journal, 14(5), 371–385.

Hamel, G., Doz, Y. L., & Prahalad, C. K. (1989). Collaborate with Your Competitors–and Win.

Harvard Business Review, 67(1), 133–139.

Hansen, J. M., McDonald, R. E., & Mitchell, R. K. (2013). Competence Resource Specialization,

Causal Ambiguity, and the Creation and Decay of Competitiveness: The Role of Marketing

Strategy in New Product Performance and Shareholder Value. Journal of the Academy of

Marketing Science, 41(3), 300–319.

Helfat, C. E., & Peteraf, M. A. (2003). The Dynamic Resource-Based View: Capability Lifecycles.

Strategic Management Journal, 24(10), 997–1010.

Henderson, R. M., & Clark, K. B. (1990). Architectural Innovation: The Reconfiguration of

Existing Product Technologies and the Failure of Established Firms. Administrative Science

Quarterly, 35(1), 9–30.

Hess, A. M., & Rothaermel, F. T. (2011). When Are Assets Complementary? Star Scientists,

Strategic Alliances, and Innovation in the Pharmaceutical Industry. Strategic Management

Journal, 32(8), 895–909.

31

Iansiti, M., & Levien, R. (2004). The Keystone Advantage: What the New Dynamics of Business

Ecosystems Mean for Strategy, Innovation, and Sustainability (Vol. 253). Boston, MA:

Harvard Business School Press.

Inkpen, A. C., & Ross, J. (2001). Why Do Some Strategic Alliances Persist Beyond Their Useful

Life? California Management Review, 44(1), 132–148.

IT-Reseller.de. (2012, August 31). SAPs Vertriebsstrategie von Business Bydesign. Retrieved May

14, 2013, from http://www.itreseller.ch/artikel/art_drucken.cfm?aid=70954

Jacobides, M. G., Knudsen, T., & Augier, M. (2006). Benefiting from Innovation: Value Creation,

Value Appropriation and the Role of Industry Architectures. Research Policy, 35(8), 1200–

1221. doi:10.1016/j.respol.2006.09.005

Jacobides, M. G., & Winter, S. G. (2005). The Co-Evolution of Capabilities and Transaction Costs:

Explaining the Institutional Structure of Production. Strategic Management Journal, 26(5),

395–413.

Katz, M. L., & Shapiro, C. (1985). Network Externalities, Competition, and Compatibility. The

American Economic Review, 75(3), 424–440.

Kawamoto, D. (2007). ByDesign, SAP Introduces On-Demand Business. CNET News. Retrieved

December 29, 2012, from http://news.cnet.com/ByDesign,-SAP-introduces-on-demand-

business/2100-1012_3-6208931.html

Koch, J. (2008). Strategic Paths and Media Management - A Path Dependency Analysis of the

German Newspaper Branch of High Quality Journalism. Schmalenbach Business Review,

60(1), 50–73.

Koch, J. (2011). Inscribed Strategies: Exploring the Organizational Nature of Strategic Lock-In.

Organization Studies, 32(3), 337–363.

Kuschinsky, N. (2008). Stabilisierung von Hersteller-Lieferantenbeziehungen als pfadabhängiger

Prozess. Frankfurt am Main: Peter Lang.

Langley, A. (1999). Strategies for Theorizing from Process Data. The Academy of Management

Review, 24(4), 691–710.

Lee, C.-H., Venkatraman, N., Tanriverdi, H., & Iyer, B. (2010). Complementarity-Based

Hypercompetition in the Software Industry: Theory and Empirical Test, 1990–2002.

Strategic Management Journal, 31(13), 1431–1456.

Lehrer, M. (2006). Two Types of Organizational Modularity: SAP, ERP Product Architecture and

the German Tipping Point in the Make/Buy Decision for IT Services. In M. Miozzo & D.

Grimshaw (Eds.), Knowledge Intensive Business Services: Organizational Forms and

National Institutions (pp. 187–204). Cheltenham, MA: Elgar.

32

Lehrer, M., & Behnam, M. (2009). Modularity vs Programmability in Design of International

Products: Beyond the Standardization-adaptation Tradeoff? European Management Journal,

27(4), 281–292.

Leimbach, T. (2007). Vom Programmierbüro zum globalen Softwareproduzenten: Die

Erfolgsfaktoren der SAP von der Gründung bis zum R/3-Boom, 1972 bis 1996. Zeitschrift

für Unternehmensgeschichte/Journal of Business History, 52(1), 33–56.

Leonard-Barton, D. (1992). Core Capabilities and Core Rigidities: A Paradox in Managing New

Product Development. Strategic Management Journal, 13(special issue), 111–125.

Levering, R., Ligthart, R., Noorderhaven, N., & Oerlemans, L. (2013). Continuity and Change in

Interorganizational Project Practices: The Dutch Shipbuilding Industry, 1950–2010.

International Journal of Project Management, 31(5), 735–747.

Li, S. X., & Rowley, T. J. (2002). Inertia and Evaluation Mechanisms in Interorganizational Partner

Selection: Syndicate Formation Among U.S. Investment Banks. The Academy of

Management Journal, 45(6), 1104–1119.

Mackie, J. L. (1965). Causes and Conditions. American Philosophical Quarterly, 2(4), 245–264.

Mallach, R. J. (2012). Pfadabhängigkeit in Geschäftsbeziehungen. Wiesbaden: Springer Gabler.

Masuch, M. (1985). Vicious Circles in Organizations. Administrative Science Quarterly, 30(1), 14–

33.

Maurer, I., & Ebers, M. (2006). Dynamics of Social Capital and Their Performance Implications:

Lessons from Biotechnology Start-ups. Administrative Science Quarterly, 51(2), 262–292.

Meyer, R. (2008). Partnering with SAP: Business Models for Software Companies. Norderstedt:

Books on Demand.

Milgrom, P., & Roberts, J. (1995). Complementarities and Fit. Strategy, Structure, and

Organizational Change in Manufacturing. Journal of Accounting and Economics, 19(2–3),

179–208.

Nohria, N., & Garcia-Pont, C. (1991). Global Strategic Linkages and Industry Structure. Strategic

Management Journal, 12(Special Issue), 105–124.

PAC. (2009). Pressemitteilungen - Das SAP Skills-Ecosystem – 160.000 Gründe, dieses Thema zu

diskutieren! Retrieved June 9, 2012, from http://www.blogspan.net/presse/das-sap-skills-

ecosystem-160000-grunde-dieses-thema-zu-diskutieren/mitteilung/42978/

Parmigiani, A., & Mitchell, W. (2009). Complementarity, Capabilities, and the Boundaries of the

Firm: The Impact of Within-firm and Interfirm Expertise on Concurrent Sourcing of

Complementary Components. Strategic Management Journal, 30(10), 1065–1091.

33

Patzelt, H., & Shepherd, D. A. (2008). The Decision to Persist with Underperforming Alliances: The

Role of Trust and Control. Journal of Management Studies, 45(7), 1217–1243.

Pitelis, C. N., & Teece, D. J. (2010). Cross-border Market Co-Creation, Dynamic Capabilities and

the Entrepreneurial Theory of the Multinational Enterprise. Industrial and Corporate

Change, 19(4), 1247–1270.

Plowman, D. A., Baker, L. T., Beck, T. E., Kulkarni, M., Solansky, S. T., & Travis, D. V. (2007).

Radical Change Accidentally: The Emergence and Amplification of Small Change. Academy

of Management Journal, 50(3), 515–543.

Porter, M., & Siggelkow, N. (2008). Contextuality Within Activity Systems and Sustainability of

Competitive Advantage. Academy of Management Perspectives, 22(2), 34–56.

Powell, W. W. (1990). Neither Market nor Hierarchy: Network Forms of Organization. In L. L.

Cummings & B. M. Staw (Eds.), Research in Organizational Behavior (Vol. 12, pp. 295–

336). Greenwich, CT: JAI.

Prahalad, C. K., & Hamel, G. (1990). The Core Competence of the Corporation. Harvard Business

Review, 68(3), 79–91.

Quinn, J. B., & Hilmer, F. G. (1994). Strategic Outsourcing. Sloan Management Review, 35(4), 43–

55.

RAAD. (2012). Wegweiser ins SAP Ökosystem | RAAD Research. Retrieved March 4, 2012, from

http://www.analyst-review.de/2011/wegweiser-ins-sap-oekosystem/

Ragin, C. C. (2000). Fuzzy-Set Social Science. University of Chicago Press.

Rothaermel, F. T. (2001). Incumbent’s Advantage Through Exploiting Complementary Assets Via

Interfirm Cooperation. Strategic Management Journal, 22(6-7), 687–699.

Rothaermel, F. T., & Boeker, W. (2008). Old Technology Meets New Technology:

Complementarities, Similarities, and Alliance Formation. Strategic Management Journal,

29(1), 47–77.

Santoro, M. D., & McGill, J. P. (2005). The Effect of Uncertainty and Asset Co-specialization on

Governance in Biotechnology Alliances. Strategic Management Journal, 26(13), 1261–

1269.

SAP. (2008a). Annual Report.

SAP. (2008b). Strategie und Ausrichtung im Mittelstand. Retrieved January 16, 2012, from

http://www.aiinformatics.com/speedforum/vortraege/01_SAP_Strategie_fuer_den_Mittelsta

nd.pdf

SAP. (2012). SAP - Partnering with SAP: SAP Services Partners. Retrieved March 14, 2012, from

http://www.sap.com/partners/partnerwithsap/services/index.epx

34

Sarker, S., Sahaym, A., & Bjorn-Andersen, N. (2012). Exploring Value Cocreation in Relationships

Between an ERP Vendor and its Partners: A Revelatory Case Study. Management

Information Systems Quarterly, 36(1), 317–338.

Schilling, M. (1999). Winning the Standards Race: Building Installed Base and the Availability of

Complementary Goods. European Management Journal, 17(3), 265–274.

Schreiner, M., Kale, P., & Corsten, D. (2009). What Really is Alliance Management Capability and

How Does It Impact Alliance Outcomes and Success? Strategic Management Journal,

30(13), 1395–1419.

Schreyögg, G., Sydow, J., & Koch, J. (2003). Organisatorische Pfade – Von der Pfadabhängigkeit

zur Pfadkreation? In G. Schreyögg & J. Sydow (Eds.), Managementforschung 13:

Strategische Prozesse und Pfade (pp. 257–294). Wiesbaden: Gabler.

Shapiro, C., & Varian, H. R. (1999). Information Rules (A Strategic Guide to the Network

Economy). Boston: Harvard Business School Press.

Simon, H. A. (1979). Rational Decision Making in Business Organization. American Economic

Review, 69(4), 493–513.

stern.de. (2012). Cloud-Computing: Die Wolke des Grauens - Digital. Retrieved July 9, 2012, from

http://www.stern.de/digital/online/cloud-computing-die-wolke-des-grauens-1795640.html

Stieglitz, N., & Heine, K. (2007). Innovations and the Role of Complementarities in a Strategic

Theory of the Firm. Strategic Management Journal, 28(1), 1–15.

Sydow, J. (2005). Managing Interfirm Networks – Towards More Reflexive Network Development?

In T. Theurl (Ed.), Economics of Interfirm Networks (pp. 217–236). Tübingen: Mohr

Siebeck.

Sydow, J., Schreyögg, G., & Koch, J. (2009). Organizational Path Dependence: Opening the Black

Box. The Academy of Management Review, 34(4), 689–709.

Sydow, J., Windeler, A., Müller-Seitz, G., & Lange, K. (2012). Path Constitution Analysis - A

Methdology for Understanding Path Dependence and Path Creation. Business Research,

5(2), 1–22.

Teece, D. J. (1986). Profiting from Technological Innovation: Implications for Integration,

Collaboration, Licensing and Public Policy. Research Policy, 15(6), 285–305.

Teece, D. J. (1992). Competition, Cooperation, and Innovation: Organizational Arrangements for

Regimes of Rapid Technological Progress. Journal of Economic Behavior & Organization,

18(1), 1–25.

Teece, D. J. (2006). Reflections on “Profiting from Innovation.” Research Policy, 35(8), 1131–

1146.

35

Tripsas, M., & Gavetti, G. (2000). Capabilities, Cognition, and Inertia: Evidence from Digital

Imaging. Strategic Management Journal, 21(10-11), 1147–1161.

Van de Ven, A. H., & Walker, G. (1984). The Dynamics of Interorganizational Coordination.

Administrative Science Quarterly, 29(4), 598–621.

Venkatraman, N., Lee, C.-H., & Iyer, B. (2008). Interconnect to Win: The Joint Effects of Business

Strategy and Network Positions on the Performance of Software Firms. In T. J. Rowley & J.

A. C. Baum (Eds.), Advances in Strategic Management (Vol. 25, pp. 391–424). Bingley

(UK): Emerald.

Weigelt, C. (2009). The Impact of Outsourcing New Technologies on Integrative Capabilities and

Performance. Strategic Management Journal, 30(6), 595–616.

Yang, H., & Tate, M. (2012). A Descriptive Literature Review and Classification of Cloud

Computing Research. Communications of the Association for Information Systems, 31(1),

34–61.

Yin, R. K. (2009). Case Study Research: Design and Methods (4th ed.). Thousand Oaks: Sage.