On the starting situation for business relationship initiation in turbulent business networks

Abstract

Purpose

The purpose of this study is to develop a model of a starting situation for relationship initiation in turbulent business networks.

Design/methodology/approach

The study is designed as an extreme single case study that takes its point of departure in a company’s bankruptcy in the Swedish automotive industry.

Findings

This study illustrates how a new business relationship can start from a resource combination previously controlled by one actor (i.e. a single company) in a turbulent business network, thereby bringing nuances to the common understanding that new relationships start in stable business networks where resource combinations are developed between actors in established business relationships.

Originality/value

Previous studies have stated that the development of a mutual orientation between actors leads to the formation of a business relationship. The business relationship then leads to resource adaptations between the two companies. The developed model, however, illustrates that this pattern can be reversed in situations of turbulence. Hence, previously adapted resources might lead to the formations of a business relationship. Based on this observation, the authors argue that there are reasons to question if previous models of business relationship initiation and development in business networks are adequately equipped for analysis in turbulent business networks.

Citation

Wadell, O. and Bengtson, A. (2024), "On the starting situation for business relationship initiation in turbulent business networks", Journal of Business & Industrial Marketing, Vol. 39 No. 13, pp. 1-14. https://doi.org/10.1108/JBIM-06-2022-0251

1. Introduction

Despite business relationships and relationship development being a much researched subject, previous research has alleged that much less is written about “how business relationships come into being” (Mandják et al., 2016, 2015, p. 137). We therefore engage in a conversation regarding the starting situation of new business relationships, and focus especially on how starting situations in network turbulence lead up to relationship initiation. A starting situation is portrayed in this paper as the situation from which the relationship comes into being, i.e. the setting that forms the basis for the start of a new relationship (Mandják et al., 2015).

The conversation about the starting situation for new relationships is important for scholars from two connected literature streams; Firstly, scholars who are interested in relationship development in general (Ford, 1980; Dwyer et al., 1987; Polonsky et al., 2010), in which the initiation is one step, i.e. the starting step, and secondly, scholars who focus specifically on the relationship initiation phase (Aaboen and Aarikka-Stenroos, 2017; La Rocca et al., 2019; Mandják et al., 2015). Regardless of the focus, social exchange theory (Blau, 1964) forms the point of departure in this conversation, as it has been used to argue that business relationships, just like social relationships, arise through minor transactions where little trust is required (Håkansson and Snehota, 1995; Håkansson et al., 2009). Such minor transactions will eventually lead to the formation of a mutual orientation between the actors, which is the core of a business relationship (Johanson and Mattsson, 1987; Johanson and Mattsson, 1992). As a business relationship is formed, the actors will gradually form actor bonds and engage in larger transactions, which, in turn, tie actors’ resources to each other through heavy investment and adaptations (Ford, 1980; Dwyer et al., 1987; Wilson, 1995). This creates interdependencies between the resources of different actors as well as strong resource combinations (Håkansson and Ford, 2002). Past research, furthermore, state that these relationship initiation processes cannot be understood in isolation; rather, they need to be understood as taking place in the context of a business network where past interactions within connected business relationships can facilitate or hinder the initiation process (Aaboen et al., 2017; Edvardsson et al., 2008; Valtakoski, 2015). Accordingly, the picture given in contemporary literature is that in a stable business network setting of long-term relationships, a mutual orientation between two actors will, through time, lead to adaptation between their resources.

However, in this paper, we want to complement this picture by adding network turbulence. Network turbulence implies a contextual setting in which stable interactions in long-term relationships are replaced by short-term interactions and frequent, and often unpredictable, exits and entries of actors (Salmi, 2000). Our focus on business network turbulence is based on empirical observations showing that actors on business markets might go bankrupt (Wadell and Åberg, 2021), close down (Havila and Medlin, 2012), become acquired (Anderson et al., 2001) and/or make relationship reactivations (Poblete and Bengtson, 2020). Such events are all characteristic of a turbulent business network (cf. Salmi, 2000). The turbulent business network is a starting situation for new relationship initiation processes that seem to be overlooked, yet are important (Waluszewski et al., 2019; Zafari et al., 2023). Previous studies have, for example, indicated that relationship-based resource adaptations can emerge earlier in network settings characterised by turbulence (Lambe et al., 2000; Zafari et al., 2023), making it interesting to expand on this finding to better understand the starting situation behind relationship initiation in turbulent networks. It has been argued that relationship initiation, as such, has been under researched (Mandják et al., 2015, 2016). Hence, an understanding of the effects of turbulence for the starting situation leading to relationship initiation is important, as it provides new nuances to our understanding of how relationships come into being.

In line with the reasoning above, the purpose of this study is to develop a model of a starting situation for relationship initiation in turbulent business networks. To do so, we make use of a single extreme case study in the Swedish automotive industry. The study takes its point of departure in the car manufacturer Saab Automobile (Saab). The case under investigation illustrates how a resource combination is developed by Saab, but eventually through network turbulence, i.e. a bankruptcy, is separated into a bankruptcy estate and a new start-up. Specifically, the case study culminates in showing how the resource combination is brought back together through the initiation of a new relationship between two new actors.

The key contribution of the paper concerns the initiation of relationships (Aaboen and Aarikka-Stenroos, 2017; La Rocca et al., 2019; Mandják et al., 2015, 2016) as it offers new insights concerning the starting situation of new business relationships. This study brings nuances to past research, as it illustrates how a business relationship can be initiated in a turbulent business network. Previous research has, based on social exchange theory (Blau, 1964), argued that the development of actor bonds (Håkansson and Snehota, 1995) leads to the formation of a business relationship, and from the initiated relationship then follows adaptations of resources between the parties in the next step (Aaboen et al., 2017; Edvardsson et al., 2008; Valtakoski, 2015). The model we develop, however, illustrates that in a turbulent business network, this pattern might be reversed so that previously adapted resources lead to the formation of a business relationship. This observation makes it possible to question if models developed to study business relationship initiation in business networks are adequate in turbulent business networks settings (cf. Zafari et al., 2023).

In the next section, the theoretical framework is introduced by presenting previous research on the starting situation of new relationship initiation and the role of resource adaptions in relationship development. Thereafter, we present or view on turbulent business networks using literature on bankruptcy, acquisitions and business relationship reactivation (Poblete and Bengtson, 2020). This part also includes a section on resource interaction (Håkansson and Waluszewski, 2002a), which gives us the possibility to elaborate on relationship starting situations based on one established resource combination within a single company (cf. Aaboen et al., 2017; Gadde et al., 2012; Håkansson et al., 2009). Then follows a methodological section addressing research design, choice of empirical case, data collection method and a description of how the analysis was conducted. After the case and case analysis have been presented, the paper ends with conclusions and implications. In this part, we outline the conceptual model, discuss contributions of the study, provide suggestions for future research and address managerial implications.

2. Theoretical framework

2.1 Previous research on starting situation of new relationship initiation

There are several models that describe how relationships develop over time (Heide, 1994; Larson, 1992; Wilson, 1995). These models commonly build on the thought that an initiation phase launches relationship development (Aaboen et al., 2017; Edvardsson et al., 2008; Mandják et al., 2015). It is argued, however, that the literature seldom presents a clear definition of the initiation process (Aaboen and Aarikka-Stenroos, 2017; Edvardsson et al., 2008; Hedaa, 1996). Edvardsson et al. (2008, p. 340) argue that “[initiation] starts when the companies in a potential relationship recognize each other and ideally ends when a business agreement is reached”. Aaboen and Aarikka-Stenroos (2017, p. 232), on the other hand, define the initiation process as “a dyadic process at the organization level, starting with awareness and ending in the agreement, or assignment that may lead to a business relationship”.

What both these definitions bring to the forefront is the idea that initiation is a process that must start from something (Ford et al., 1998; Håkansson and Snehota, 1995). Edvardsson et al. (2008), for example, recognise that before initiation, there is an “unrecognised status” where actors “don’t know each other”. They are not alone in this position. In fact, many models of relationship development and initiation start from a similar assumption. Ford et al. (1998, p. 30) call the time before a relationship is formed the “pre-existing starting situation”, while Hedaa (1996) terms it the “pre-awareness phase”, which corresponds, for example, to Dwyer’s (1987) view of initiation starting with recognition and awareness. Mandják et al. (2015, p. 37), whose terminology is used in this paper, refer to it simply as the “starting situation”. It is argued that, in the starting situation, companies are not only unacquainted with one another (Edvardsson et al., 2008; Mandják et al., 2015; Valtakoski, 2015), but they also lack a common relationship history and have no past shared experiences or knowledge about each other (Ford, 1980; Polonsky et al., 2010). A reason for this view of the starting situation is historical, as initiations in early studies on relationships were viewed to develop from traditional market settings of arms-length relationships (Dwyer et al., 1987; Easton, 1992; Ford, 1980).

However, more recent studies of business relationship initiation (Aaboen et al., 2017; La Rocca et al., 2019; Aarikka-Stenroos and Halinen, 2007) have taken a network perspective and acknowledge that relationship development needs to be considered as embedded in a larger structure (Aaboen and Aarikka-Stenroos, 2017; Edvardsson et al., 2008; Valtakoski, 2015). We know from past research that two actors who want to develop a relationship will not do this in a “vacuum” (Aaboen et al., 2017), but that the initiation might be impacted by a third actor who acts as a mediator, or through individuals’ social contacts (Aaboen et al., 2017; Aarikka-Stenroos and Halinen, 2007; Ring and Van de ven, 1994).

A business network is a structure where the process of interaction between actors over time lead to the creation of connected long-term business relationships (Ford et al., 1986; Ford and Håkansson, 2006; Håkansson and Ford, 2002). The connectedness of relationships implies that they cannot be fully understood in isolation as they impact each other (Anderson et al., 1994; Håkansson and Snehota, 1995; Snehota, 1990). Interactions between actors over time mean that they, in their struggle for economical, technological and social gains, connect their activities and resources to certain patterns and constellations (Håknasson and Johanson, 1992; Håkansson and Snehota, 1995). Three interrelated networks of actors, activities and resources are described to form the underlying structure of the business network (Harrison and Håkansson, 2006; Wedin, 2001). The activity network relates to the linking of activities of the actors, the resources network relates to how actors’ resources might become more or less adapted to each other, whereas the actor layer relates to how individuals through interaction develop trust, commitment and a mutual orientation (Håkansson et al., 2009). The heavy investment made also implies that networks are perceived as rather stable structures, often characterised by incremental changes (Halinen et al., 1999).

2.1.1 Previous views on resources adaptations in business relationships

Despite the fact that there are several studies which acknowledge that business relationship initiation takes place in a business network, it is still often assumed that factors such as resource adaptations “cannot influence relationship initiation” (Valtakoski, 2015, p. 108; Wilson, 1995). The relationship initiation is instead assumed to start from awareness, trust building and other bond-related aspects, which are perceived as needed before resources are adapted. As pointed out by Johanson and Mattsson (1992, p. 207), “through the exchange relationship the actors learn about each other and develop some trust in each other. On that basis, they adapt and develop their resources […]”.

In literature that views relationships development in both traditional market situations (Dwyer et al., 1987) and in business networks (Aaboen and Aarikka-Stenroos, 2017; Edvardsson et al., 2008; Valtakoski, 2015), assumptions about relationship initiation are based on social exchange theory (Blau, 1964) and thus perceived to develop gradually. This implies starting with a large distance and small transactions and gradually, over time, actors become familiar with each other and start trusting each other and will then move closer to each other, making mutual adaptation and investments in each other’s resources (Dwyer et al., 1987; Ford, 1980 Valtakoski, 2015).

An adaptation is a specific investment in another actor’s resources, such as machinery, manufacturing technology or specific knowhow (Brennan and Turnbull, 1999; Cannon and Perreault, 1999, p. 443–444; Hallén et al., 1991). Models of relationship development (Ford, 1980), as well as studies of relationship initiation (Valtakoski, 2015), assume that all adaptations and resource ties (interdependencies between resources) are created in the later stages of a relationship (see Table 1 below). For example, Ford (1980) describes a large technological distance in the initiation of relationships, with adaptations appearing first in the “long-term stage”, which is the fourth of five stages of their model. Dwyer et al. (1987, p. 19) argue that it is not until the commitment phase, the fourth phase before dissolution, that companies are tied to each other with transaction-specific investments, or what in this paper is referred to as adaptations (cf. Cannon and Perreault, 1999, p. 444). This assumption is also expressed in later stage models (Batonda and Perry, 2003; Polonsky et al., 2010). Batonda and Perry (2003), for example, point out that adaptations between the parties arise in the maintenance stage, which is the fourth of five stages in their model. Moreover, Polonsky et al. (2010) place adaptation in later stages of relationship development when relationships have become more stable and durable over time.

The different views on the starting situation for business relationships, as described in previous literature, are presented in Table 1 below. This table summarises literature on relationship initiation and development, focusing not merely on their view of the starting situation, but also on their view of the point in time when resource adaptations are made. In the next section, this starting situation is elaborated on further by the introduction of turbulent business networks.

2.2 Starting situation in turbulent business network

In this paper, network turbulence is interpreted as business network settings that are less tightly structured due to shifts and changes of actors in the network and a situation in which there is an actor focus on finding new business partners (Salmi, 2000). Despite limited knowledge of how relationships are initiated in turbulent business networks (Zafari et al., 2023), we know more about such networks’ general characteristics (Salmi, 2000; Hurmelinna, 2018). Past research has shown that a turbulent business network (Salmi, 2000) is characterised by radical changes, rather than incremental ditto (Halinen et al., 1999). In turbulent business networks, there are frequent exits and entries of actors (Salmi, 2000). Reasons for exiting a network might be closure (Havila and Medlin, 2012) or bankruptcy (Wadell et al., 2019), and the reason for joining might be mergers or acquisitions (Anderson et al., 2001). Such circumstances then imply that resources due to different trigger issues (Mandják et al., 2015) are shifted and moved around between different actors.

Past research has shown that besides long-term business relationships (Johanson and Mattsson, 1987) where actors have adapted their resources to each other (Hallén et al., 1991), there can be many relationships that are dormant or resting (Zafari et al., 2023) in turbulent networks settings. Even though relationships have been triggered to end and have become dormant, companies may maintain not only social bonds (Gidhagen and Havila, 2016; Havila and Wilkinson, 2002; Pressey and Mathews, 2003) but also mutual adaptations and investments (Hurmelinna, 2018). From the literature, it seems that the main reason behind keeping social and customised physical resources from ended relationships is the possibility for future business relationship reactivation (Poblete, 2017; Polonsky et al., 2010; Zafari et al., 2023).

Reactivation of an old relationship has been pointed out as a less costly and fast alternative to develop a new relationship (Pressey and Mathews, 2003). For successful reactivation, there are a number of factors to consider. Poblete and Bengtson (2021) claim, for example, that boundary spanners can be important as they keep contacts well informed regarding potential resources in the environment. Hurmelinna (2018) states that the most common reason to return to a business partner is previous resource adaptations left between actors. Poblete and Bengtson (2020, p. 10) even point out that reactivation was considered by an actor because of “[…] a number of adaptations resulting in resource ties between the parties, which made it interesting from an economical point of view to return to the relationship”.

The reason behind actors wanting to return to old relationships with previous resource adaptations is based on resources heterogeneity (Hägg and Johanson, 1982; Penrose, 1959) i.e. that a resource’s economic value spurs from how it is combined with other resources (Håkansson and Waluszewski, 2002a; Håkansson et al., 2009). Resources heterogeneity can have different roles to play at different stages of relationship development:

• In turbulent business networks and in a situation of reactivation, the heterogeneity of resources and their lack of value in isolation drives actors to return to a counterpart if there are previous adaptations between their resources.

• In stable long-term relationships, resource heterogeneity plays an important role in holding the relationship together (Håkansson and Snehota, 1995), and finally.

• In the starting situation of new relationships, resources heterogeneity is an important assumption as actors develop business relationships to draw benefits from mutual resource adaptations (Håkansson et al., 2009).

What all these three perspectives on heterogeneity, then, have in common is that they all start from the assumption that resources heterogeneity is developed by actors in business relationships.

However, viewing resource constellations in a network setting, it has been shown in previous research that there are, of course, resources adaptations also between resources, i.e. in resources combinations, outside established business relationships, for example, within a single company (Wedin, 2001) or among several actors (Håkansson et al., 2009). The existence of such resources combinations then opens up an opportunity to consider a new possible starting situation for relationship initiation in turbulent business networks. Such a starting situation might spur from the notion that a resource combination of heterogeneous resources (i.e. resources in one company) is separated in a turbulent business network. Given that resources within a company are adapted to each other, there can, in situations of turbulence, be shifts in resources control (e.g. through a bankruptcy or de-merger), indicating that an actor without a common past relationship will control resources that are best used in combination.

2.2.1 Towards view on starting situations based on resource interaction

To comprehend this type of starting situation, we cannot use literature that starts in the interaction and business relationship between actors (Ford et al., 1986; Ford and Håkansson, 2006; Johanson and Mattsson, 1987). To comprehend a starting situation of the kind, it seems that we need, instead, to start out from an analytical perspective that has a more specific focus on the resource dimension of business networks. Håkansson and Waluszewski (2002a) have developed a resources interaction approach that offers such a perspective. This approach implies that interaction does not merely take place between actors in business relationships, but also directly between heterogonous resources within and between companies (Håkansson et al., 2009). From this point of view, a resource is perceived as an object “with actual or potential use within a certain network context where it can be combined with other resources” (Baraldi et al., 2012a, p. 271).

Håkansson and Waluszewski (2002a), furthermore, maintain that it is possible to divide resources into two groups, in particular: organisational (relationships and business units) and technical (facilities and products). In contrast to the relationship initiation and development literature (Aaboen et al., 2017; Mandják et al., 2015; Polonsky et al., 2010) Håkansson and Waluszewski (2002a) do not study relationships as ends, but as central means for connecting actors and their internal resources to each other. Business units can, in this context, be understood as “resources which incorporate the knowledge, identity, and reputation of an organization” (Baraldi et al., 2012, p. 268). Business units are the “active actors” (Wedin, 2001, p. 41) in this model, i.e. the ones that have agency (Håkansson and Waluszewski, 2002a). Business units, thus, hold the knowledge of how resources should be used (Snehota, 1990). Technical resources, in terms of products and facilities, are also given an extended role in the resource interaction approach (Håkansson and Waluszewski, 2002a).

Through resource interaction, combinations of these resources emerge. However, such combinations are not pre-existing but built up through gradual investments (Gadde et al., 2003). Through such investments, different resources within and between business units become adapted (Hallén et al., 1991) and tied (they become interdependent) to each other (Baraldi et al., 2012). Together, combinations of organisational and technical resources become investment-heavy structures that are costly to break (Håkansson and Waluszewski, 2002a). Regardless of the type of resource, the heaviness (large investments in adaptations) of two resources “forges them together” (Håkansson and Waluszewski, 2002a, p. 226), hinders them from being moved (Håkansson et al., 2009), and establishes a particular investment logic or path in a business network (Håkansson and Waluszewski, 2002b). Such paths can either be positive for business as past investments create a platform for future exchange (Håkansson and Ford, 2002) or act as a negative barrier that holds back change (Gadde and Mattsson, 1987).

Accordingly, it is likely that if such a resource combination is brought apart in a turbulent business network, actors gaining control of resources from separated resource combinations will try to bring them back together. Such a starting situation for the initiation of new business relationships is not much elaborated in previous literature (Aaboen et al., 2017). Hence, we will address this here by the means of a single case study, as further described in the next section.

3. Methodology and research design

This study is part of a larger research project on the empirical phenomenon of bankruptcy and its network effects. We have previously published studies based on a bankruptcy estates work to attract supplier resources (Wadell et al., 2019) and an acquirer’s work to reactivate a business relationship following a takeover of a bankrupt company (Wadell and Åberg, 2021). Collecting and coding the empirical material, the issue of relationship initiation and the possibility to nuance the starting situation came up as interesting aspects that we were able to shed new or additional light on, through our research. Our choice of methodology and design for this particular study was subsequently guided by an attempt to develop a model of a starting situation for relationship initiation. As business markets are viewed as constantly changing (Halinen et al., 2012), a process perspective is applied (Langley, 1999; Pettigrew, 1997; Van de Ven, 2007). In line with Pettigrew (1997, p. 338), process is understood as “a sequence of individual and collective events, actions and activities unfolding over time in a context”. To conduct the process study, we have chosen to use a single case study approach. Single case studies are commonly seen as a great tool for developing new theory (Halinen and Törnroos, 2005; Piekkari et al., 2010; Siggelkow, 2007).

3.1 Case selection

To explore how a new relationship can be initiated in a turbulent business network, we decided to select an extreme case (Seawright, 2016). The study of extreme cases can be advantageous when little is known about a phenomenon (e.g. the starting situation of a relationship initiation) (Seawright and Gerring, 2008). The specific case chosen is extreme because of the heavy adaptations that exist between a resource combination within a single company. Despite the company’s bankruptcy, the heavy adaptations eventually forced the resource combination back together.

The selected case concerns the bankruptcy of the iconic Swedish automotive manufacturer Saab Automobile (Saab). The Saab case has, as mentioned, been discussed in other studies, but the data for this particular case has not been used previously, as it is a new angle on the phenomenon of bankruptcy and its network effects. The automotive industry has a long history, as appropriate for business studies (Chandler, 1964), and Sweden has been shown to have a particularly open climate for research (Mattsson and Johanson, 2006). In considering Saab in this paper, we have focused particularly on the starting situation behind a relationship that was initiated between Saab’s bankruptcy estate and the new start-up LeanNova after Saab went bust in December 2011. Saab had a world-class R&D department; as they went bankrupt, parts of their former R&D management, together with the science park Innovatum and the governmental investment fund Fouriertransform, worked to keep the competence together in the new company LeanNova. Saab’s technical resources, on the other hand, ended up in the control of Saab’s bankruptcy estate.

The case was chosen in order for us to be able to problematise some of the assumptions in previous literature on the starting situation for relationship initiation. As explained by Aaboen et al. (2017, p. 36), in a book on the initiation of business relationships in start-ups, “In most models, the initiation process is usually assumed to have taken place when the interaction between resource entities occurs”.

3.2 Collection of empirical material

The empirical material for the study has primarily been collected by the authors through 20 in-depth interviews, ranging in length from 30 min to 1 h and 45 min (see Table 2 below). The interviewees were mainly identified through “snowball-sampling” (Miles and Huberman, 1994). Interviews were conducted between November 2016 and August 2019 and were recorded and transcribed. Some of Saab’s and their former suppliers’ facilities have been visited during the interviews. These visits were important for the development of the thick case description (Lincoln and Guba, 1985).

The interviewees were selected from Saab, Saab’s bankruptcy estate, Innovatum and Fouriertransform. It was determined that they all had valuable information regarding the process under study. The questions posed to the interviewees have focused on letting them describe their perspective on the unfolding process (Langley et al., 2013), and the interviewee protocol has been revised continually (Gioia et al., 2013).

To complement the interviews collected, and to triangulate the interviewee’s information further (Denzin, 1978), a large set of secondary sources has been used. The secondary sources include newspaper articles, research articles, books about Saab, internal documents from the science park Innovatum, public documents from Fouriertransform, as well as documents from Saab’s bankruptcy estate. When studying these materials, the sources and the authors were critically scrutinised (Bowen, 2009).

3.3 Analytic process

In line with many case studies on relationships and business networks, the analytic process has been one of systematic combining, using an abductive logic (Dubois and Gadde, 2002, 2014). Thus, the analytical process has implied a constant movement back and forth between empirical findings and theory. In terms of analytic strategy, we have followed the advice of Langley (1999) and Gehman et al. (2018) to combine different analytic strategies to circumvent their individual pitfalls (Langley, 1999). Langley (1999) presents seven strategies for theory creation, of which we have combined visual mapping and temporal bracketing for this study.

The analytic work started already during the interviews, in which a constant effort was made to see patterns and similarities in the different interviews. Consequently, much of the initial work was related to figuring out what this case was really about (i.e. casing) (Ragin, 1992). The work to see patterns and similarities later involved using interview transcripts and documents to write down events over time (Van de Ven, 2007) into a visual map, constantly comparing this visual map to literature on business relationship development and initiation.

As we started to see different episodes in the material, we used temporal bracketing to delineate further what was going on in the case. The empirical case and theoretical framework were then aligned with this new structure. This includes a clear element of creativity (Klag and Langley, 2013; Langley, 1999) as well as hermeneutical reading and re-reading of the material (Einola and Alvesson, 2019). The outcome of this process then formed the base for the structure of the paper.

4. Case: developing, keeping and bringing resources back together

Below, the empirical case is presented. It starts in a resource combination (i.e. its establishment) and follows it as it is separated until it is brought back together again.

4.1 Saab: bundle of physical and human resources

In February 2010, the Dutch sports car manufacturer Spyker acquired Saab. At the time, Saab was in control of one of Europe’s most modern car manufacturing plants. Located in Trollhättan, in western Sweden, the plant was about 480,000 square metres in size and had an annual production capacity of about 190,000 cars. The American automotive giant General Motors (GM) had previously owned the plant and had, over a period of 20 years, invested millions of Euros to make modernisations and increase the plant’s efficiency. GM’s investments were not free of charge for other actors, and the Swedish government had also invested heavily in the region, contributing to new infrastructure and the development of a science park environment closely connected to Saab.

Saab was also the owner of several cars that were in production or under development. The most prominent development project was probably the Phoenix-architecture, expected to become the foundation of Saab’s future car models. Saab’s infrastructure also included computers, terminals, laboratories, machinery for product development and patents, including design as well as domain rights to different technical components.

Saab controlled not only the physical assets but also a world-class R&D department. Even with cuts to the department made by GM in previous years, it still included about 1,000 engineers, many of whom had experience working with one of the world’s largest automotive manufacturers because of Saab’s integration with GM. The engineering department had a wide range of capabilities regarding car manufacturing, including areas such as body, chassis, propulsion and electrics. Furthermore, the R&D department also had the ability to coordinate different – and highly complex – internal and external units to build complete cars. Over time, the engineers’ work had become strongly integrated with the physical assets mentioned above. One interviewee even compared owning only Saab’s physical assets without employing the engineers to having a CD full of information, but not being able to use it.

Despite the competence of its R&D department, under Spyker’s ownership, Saab could not afford to keep it all in-house. During the period of Spyker’s ownership, Saab started to embark on a number of spinoffs. The spinoffs were done with support from the local science park Innovatum and the governmental investment fund Fouriertransform. These spinoffs allowed Saab to reduce their cost burden while keeping access to resources by forming business relationships with the spinoffs.

The spinoffs allowed Saab to keep the R&D department’s core competence – taking on, coordinating and building complete cars – in-house. Meanwhile, Saab tried to find ways to become profitable, but the burden of the financial crisis of 2008–2009 would prove too great; during 2011, Saab came closer and closer to bankruptcy. Despite mounting problems, only a small fraction of the employees at the R&D department left to work for competitors, but in the end it was simply not possible to keep Saab going. On 19 December 2011, after having manufactured more than 4.5 million cars in more than 60 years, Saab was declared bankrupt.

4.2 Keeping the assets of the bundle

Under Swedish law, Saab’s physical assets, with an estimated value of 173m euro, now became the property of a bankruptcy estate. The bankruptcy estate was managed by two court-appointed trustees from two different law firms that, over the following couple of months, hired about 250 employees to manage the task. The trustees and their employees were charged with selling the bankruptcy estate as profitably as possible and redistributing the proceeds to Saab’s creditors.

To be able to sell Saab’s assets to the highest bidder, it was critical for the bankruptcy estate to understand the different properties that it owned. This included written material explaining the function of different technical components that had been developed by the R&D department. With this as a background, soon after the bankruptcy petition, they started a process by locating written material that would be of interest for potential new owners. All the written material was compiled in a document that was to serve as a basis for answering questions from potential acquirers.

However, as the bankruptcy estate started the process of selling, they were flooded with questions regarding the technical components in the cars and Saab’s former car architecture, the Phoenix. One of the companies interested in Saab’s assets was the potential acquirer Mango (fictive name). Mango made extensive demands concerning access to specific information about some of the technical equipment, in particular information regarding Saab’s Phoenix architecture. At the beginning of February 2012, Mango even expressed their dissatisfaction with the information given by the bankruptcy estate in the document they had compiled, and made it clear that they were not willing to be involved in the acquisition process without a more detailed run-through of the Phoenix architecture.

4.3 A Need to bring the bundle back together

The problem for the bankruptcy estate was that at the end of December 2011, they had been forced to dismiss Saab’s approximately 3,000 employees, including roughly 1,000 engineers. To do the work that Mango, as well as other acquirers and stakeholders, required, the bankruptcy estate needed the engineers who had been working with Saab’s R&D. If they could not get those specific people, it would likely take months for new engineers to get to know the material and be able to present it to sellers. Such an additional time delay would incur heavy costs and be problematic for the bankruptcy estate.

Following the bankruptcy, to support Saab’s new owners during the transition period, Saab’s former President of Product Development had been hired by the bankruptcy estate and the VP of Product Development by the science park Innovatum. In the face of the challenges outlined above, the bankruptcy estate turned to these former managers and their new organisations for advice. The bankruptcy estate was then advised to turn to the newly started engineering consultancy company LeanNova.

To retain the competence needed to coordinate larger automotive projects and build complete cars, a number of industrial actors, including former Saab R&D managers, the science park Innovatum and the governmental investment fund Fouriertransform undertook a joint effort to create an engineering consultancy company that could rent out their knowledge when needed. The resulting company, LeanNova, was officially established on 1 February 2012. From its inception, it employed some of the most highly skilled engineers from Saab.

In LeanNova’s business plan, the founders had calculated with the bankruptcy estate as well as a potential acquirer of Saab needing to purchase engineering services from them. This being the case, it was no surprise to LeanNova when the bankruptcy estate called them at the beginning of February 2012. The consultancy firm saw this potential cooperation as a great opportunity for their new business, and for the community of Trollhättan to recover from the loss of Saab. The bankruptcy estate sent over an initial specification regarding what they needed help with, and, shortly thereafter, a first meeting was held between them and LeanNova.

4.4 Becoming partners

LeanNova became extremely valuable for the bankruptcy estate. After a short negotiation, they settled an internal agreement to do business with the bankruptcy estate. LeanNova’s first objective was to help the bankruptcy estate with the potential acquirer Mango. Mango was very satisfied with LeanNova’s work and, following this success, the bankruptcy estate hired LeanNova for a wide range of other tasks. The relationship between the bankruptcy estate and LeanNova continued until the summer of 2012 and included frequent meetings, negotiations and agreements. Both sides described the relationship as beneficial. In contrast to many other suppliers, LeanNova was willing to trust the bankruptcy estate and to deliver their engineering service before they got payments. The payments from the bankruptcy estate were made with about 30 days delay, which was less than the industry praxis of 60 days. The bankruptcy estate’s prompt payments became important for LeanNova to get cash-flow in their new business.

During the summer of 2012, the bankruptcy estate was acquired by the Swedish automotive manufacturer National Electric Vehicle Sweden (NEVS), with its Chinese ownership structure. Much like the bankruptcy estate, NEVS was in desperate need of engineering consultancy; thus, after the acquisition, they became heavily involved with LeanNova. After just a couple of months, LeanNova had approximately 200 employees working for NEVS. In the beginning of December 2013, about 17 months after the acquisition of Saab’s assets, NEVS started the production of cars in the Trollhättan plant again.

5. Analysis

The investigated case illustrates how a resource combination in a business network that becomes turbulent is established, separated and eventually brought back together in a new business relationship. This is further analysed below.

5.1 From one company: establishing and separating resources in business network

The case illustrates that the actor Saab operates in a business network (Håkansson and Snehota, 1995) and in this, it had built what Håkansson and Waluszewski (2002a) have termed a “heavy resource combination”. This cannot be done overnight, but must be built through gradual investments over time (Håkansson and Waluszewski (2002b). As pointed out, both Saab’s owners and a number of external actors in the Saab business had invested considerable amounts of money in building the resource combination. Saab, on the one hand, had built up technical resources (Baraldi et al., 2012; Håkansson and Waluszewski, 2002a) in terms of laboratories, terminals, computers and products under development, such as the Phoenix architecture. On the other hand, they had also established the organisational resource (Baraldi et al., 2012; Håkansson and Waluszewski, 2002a) of Saab’s R&D business unit. One indication of the value of the resource combination is the fact that Saab’s physical assets alone were estimated to have a monetary value of approximately 173m Euros.

There existed considerable adaptations (Baraldi et al., 2012; Hallén et al., 1991) between the business unit and the technical resources of Saab. These adaptations included the special competence of the R&D department that was necessary to manage Saab’s technical resources. Consequently, the knowledge of employees who had previously developed the laboratories and the products was crucial to operate them (cf. Snehota, 1990). As pointed out by Baraldi et al. (2012), adaptations between organisational and technical resources made it both costly and time-consuming to break the resource combination.

The efforts undertaken by Saab and other actors to keep the resource combination together, despite financial problems, further indicate how their large investments had left them unwilling to separate this heavy resource combination (Håkansson and Waluszewski, 2002a). There was, thus, a particular investment logic or path established (Håkansson and Waluszewski, 2002b, p. 562). As their financial difficulties grew, to stay on this path, Saab tried to keep access to resources by spinning off parts of the R&D department but maintaining control through new relationships. Eventually, however, these efforts were not sustainable, as the separation of the resource combination was eventually triggered (Mandják et al., 2015) by Saab filing for bankruptcy on 19 December 2011. This filing then implied turbulence in the business network (Salmi, 2000).

5.2 Resource-keeping during separation in turbulent business networks

The case at hand illustrates that the bankruptcy triggered (Mandják et al., 2015) the separation of Saab’s resources into, what according to Håkansson and Waluszewski (2002a) can be seen as, two new business units (actors). Saab’s former technical resources came under the control of one actor, the bankruptcy estate, while the competence of Saab’s R&D business unit was kept by the formation of the start-up LeanNova. Thus, as pointed out by Salmi (2000), turbulence in a business network imply an exit by actors but also that new actors make an entrance.

In line with common business logic (Håkansson and Snehota, 1995), the case then illustrates that for the possibility of a new relationship to develop from a resource combination within one company, actors need to find economic value in keeping the resources together. This finding corresponds well with what is described in the literature on relationship reactivation, where it is pointed out that after investments are separated, they are commonly brought back together because of the perceived value of the resource combination (Polonsky et al., 2010; Poblete and Bengtson, 2020).

It seems that the bankruptcy estate kept the technical resources without an initial awareness (Dwyer et al., 1987) that their technical resources needed the R&D business unit’s knowledge to be used (Snehota, 1990). As the case with the bankruptcy estate illustrates, under Swedish law, resources are kept to provide benefits for creditors when selling them. As in the case of a reactivation (e.g. Poblete and Bengtson, 2020), the bankruptcy estate realised that to do this, they needed to bring resources that had been combined in the past back together.

On the other hand, the R&D business unit was kept together with an awareness (Dwyer et al., 1987) that the new owner of Saab’s technical resources would need their support because of the resource adaptations. Even from the outset, the founders of LeanNova made economic calculations that certain amounts of revenue would come from the bankruptcy estate and later from the new owner of Saab’s technical resources. Thus, it is shown that a starting situation for a new relationship in this type of circumstance can be established with the awareness (Dwyer et al., 1987) that others in the future will need the scarce resources in one’s own control.

The case further illustrates that a new relationship that develops from a resource combination in a turbulent business network, like any relationship (Håkansson and Snehota, 1995), is not established by one actor alone (Aaboen et al., 2017; La Rocca et al., 2019). It takes at least two actors that find it hard to disregard past investments (Gadde and Mattsson, 1987) and thus during turbulence (Zafari et al., 2023) decide to keep resources as a potential platform for future interaction (Håkansson and Ford, 2002).

5.3 Initiation of new relationship and reactivation of resources

The relationship could not be initiated without actors’ activities to move the relationship from what Edvardsson et al. (2008) call an “unrecognized to recognized” status. In the case, we can see that the customer (the bankruptcy estate) moves the supplier (LeanNova) from unrecognised to recognised. This is done when the customer becomes aware (Dwyer et al., 1987) of its need, and that LeanNova can solve it. However, the supplier had already recognised (Edvardsson et al., 2008) the customer. For the supplier, the change of status (Edvardsson et al., 2008) is, thus, rather a move from an expectant status to an expected status. The supplier expected the customer to contact them and initiate a relationship – and this was also the outcome. Based on their knowledge of Saab’s resources, they were not surprised by the contact; however, they could not predict exactly when the customer’s need for their competence would surface.

In the case at hand, the trigger issue (Mandják et al., 2015) that initiated the customer’s search for a supplier was the threat to the value of its resources because of the resources’ separation from their adaptations. In order not to lose the value of the resources in their possession, the customer needed help from the supplier (cf. Gadde and Mattsson, 1987).

In this situation, the former President of Saab’s Product Development became important as a boundary spanner (Poblete and Bengtson, 2021) to initiate the relationship. For a starting situation of a new relationship in a turbulent business network (Salmi, 2000) to be established, it is therefore possible that the knowledge of employees becomes essential (Snehota, 1990). In such circumstances, relationships with other actors in the business network (Aaboen et al., 2017; Anderson et al., 1994) are also important. In the case of Saab, this is illustrated by the connection with the local science park Innovatum and Saab’s former VP, who was now employed there. These findings are in line with past research in highlighting the role of previous social relationships (Mandják et al., 2015; Valtakoski, 2015; Edvardsson et al., 2008) and involvement of other business network actors (Aarikka-Stenroos and Halinen, 2007) as important in the initiation of relationships.

The relationship is initiated by the bankruptcy estate’s attraction (Dwyer et al., 1987; Aaboen and Aarikka-Stenroos, 2017) to LeanNova, whose business unit is heavily adapted to their technical resources. In this case, the resource adaptations are not developed only in the later stages of the relationship development process, as argued in previous research, to be the only possibility (Batonda and Perry, 2003; Dwyer et al., 1987; Ford, 1980; Polonsky et al., 2010; Valtakoski, 2015); rather, they are present in the starting situation (Mandják et al., 2015) of a new relationship. The already existing adaptations between the resources controlled by the actors make it hard to argue that the relationship started in what resembles a traditional market, with an arm’s length (Larson, 1992) relationship that is simple to terminate (Dwyer et al., 1987).

As the case illustrates, the bankruptcy estate was short of money and time. It would have been possible to search for an alternative supplier in the turbulent business network (Salmi, result in 2000). However, even if other actors had been available, it would have taken them months just to get familiar with Saab’s technical resources. LeanNova could start working immediately because of their previous adaptations, meeting the bankruptcy estate’s needs much quicker than any possible competitor. Along the lines of Håkansson and Waluszewski (2002a, p. 226), it can thus be argued that the heaviness of two resources forges them together again. The burden of past investments becomes visible as it hinders the bankruptcy estate from developing their technical resources in different directions from the established path (Håkansson and Waluszewski, 2002b). LeanNova, in turn, seemed to be aware of this path and that the bankruptcy estate would eventually need their help. LeanNova was also committed to the relationship, seeing it as an opportunity to get fast access to the resources (Håkansson and Snehota, 1995; Hurmelinna, 2018).

LeanNova was willing to provide the bankruptcy estate with the help it needed and embarked upon an episodic- (Halinen and Tähtinen, 2002) or interimistic (Lambe et al., 2000) relationship with them. In contrast to other suppliers, LeanNova even trusted the bankruptcy estate enough to deliver engineering services before they were paid. Accordingly, this case also demonstrates that a starting situation can arise from actors that control resources seeing mutual future benefits when engaging in business activities (La Rocca et al., 2019) and reactivating resource combinations (Poblete and Bengtson, 2021).

6. Conclusion and implications

In this final section, we will outline our model of a starting situation for relationship initiation in turbulent business networks. We will also discuss the findings in connection to the current view of the starting situation of new relationships.

6.1 Conceptual model and conclusions

Figure 1 illustrates our model of a starting situation for relationship initiation in turbulent business networks. The model’s point of departure is that markets are viewed as business networks that can become turbulent (Salmi, 2000). More specifically, three interrelated networks (actor, activity and resources) form the underling structure of the business network (Håkansson and Snehota, 1995; Håkansson and Waluszewski, 2002a; Håkansson et al., 2009). In past literature on relationship initiation and development, business relationships have been considered to develop in a situation where there is an interaction between actors that leads to formation of resources combinations (i.e. starting in the actor network) (Aaboen et al., 2017; Ford, 1980; Valtakoski, 2015).

The starting point for our model is different, as its point of departure is the interaction between resources in a single company (i.e. starting in the resource network). Accordingly, at the start of our model, a relationship is viewed only as a resource among others in different resource combinations (Håkansson and Waluszewski, 2002a).

Resource combinations of this sort can include any set of organisational (relationships and business units) and technical (product and facility) resources. Regardless of combination, it is established through heavy investments by one or several actors in a business network (Håkansson and Waluszewski, 2002a). The combination was developed through constant combining of resources (Baraldi et al., 2012) in an ongoing process (illustrated by arrows in the figure). Through this process, resources are adapted to each other, and result in one heavy combination of different interdependent resources (cf. Håkansson and Waluszewski, 2002a).

In the model, a resource separation trigger eventually breaks this resource combination apart. This trigger can be a bankruptcy, as in the case examined in this paper, but there are other types of events, such as e.g. closures (Havila and Medlin, 2012) or acquisitions (Hurmelinna, 2018). Regardless of the type, such a trigger is an event that acts as a catalyst for change, which prompts the separation of resources (Mandják et al., 2015, p. 36).

The model continues after the separation to illustrate that previously combined resources (e.g. facility X, product Y and business unit Z) end up in disconnected business units (i.e. actor 1 and 2). During the time that the resource combination is separated, these business units exist in a turbulent business network (Salmi, 2000). There are, thus, actors that are leaving the network and others that are entering (Salmi, 2000). In this situation, the new actors who enter the turbulent network (Actor 1 and Actor 2) must work to maintain the resources that they have taken control over. This dormant time period will eventually end when an actor controlling a separated part of the combination sees an opportunity to start using the resources in combination again. At this point, the model’s analytical focus shifts from a focus on the interaction between resources (Håkansson and Waluszewski, 2002a) to the interaction between actors (Ford et al., 1986).

When an actor attempts to start using its separate parts of the combination, this will act as a resource reactivation trigger. A resource reactivation trigger is an event that acts as a catalyst for business units to try to get the resources combination back together. The driving force behind this process is the interdependence between two different actors’ heterogonous resources. This interdependence among heterogeneous resources (Penrose, 1959) has been created through past adaptations (Håkansson and Waluszewski, 2002a) and leads to the formation of a new relationship (i.e. a mutual orientation).

The model accordingly starts in the resource dimension (a resource combination controlled by one company) of a business network that becomes turbulent (Salmi, 2000). It is first in the next step, as interactions between actors (Håkansson and Snehota, 1995) start, that the actor dimension becomes in focus.

6.2 Contributions of the study

The purpose of this study has been to develop a model of a starting situation for relationship initiation in turbulent business networks. By doing so. we intended to add knowledge to our contemporary understanding of how business relationships are initiated. The model contributes to past literature on relationship development (Ford, 1980; Polonsky et al., 2010), initiation (Aaboen et al., 2017; Edvardsson et al., 2008; La Rocca et al., 2019) and interaction in business networks (Håkansson et al., 2009) in three ways in particular.

Firstly, in much of the past research, business relationships have been considered to develop in stable business networks (Håkansson and Snehota, 1995) where there are resources combinations adapted through time between actors in long-term business relationships (Aaboen et al., 2017; Ford, 1980; Valtakoski, 2015). Our model, instead, contributes by illustrating a starting situation where business relationships can be initiated from one resource combination (Håkansson and Waluszewski, 2002a, 2002b) previously controlled by a single company in a turbulent business network. In line with this finding, we argue that the logic behind the initiation of a new relationship cannot always be understood from previous interaction between actors in a business network (Aaboen et al., 2017; Ford, 1980; Valtakoski, 2015). In accordance with Aaboen et al. (2017, p. 36), our findings illustrate that initiation and its starting situation at times are better understood through an in-depth analysis of the path dependence of previous resource interactions. Such analysis can then start within, but also between, companies (Håkansson et al., 2009; Håkansson and Waluszewski, 2002a) in a turbulent business network (Salmi, 2000).

Secondly, in past research on relationship initiation (Aaboen et al., 2017; Edvardsson et al., 2008; La Rocca et al., 2019) and development (Dwyer et al., 1987; Ford, 1980), it has been assumed that business relationships can be perceived as a gradual process where social exchange (Blau, 1964) leads to the creation of a bond of mutual orientation (Johanson and Mattsson, 1987) between actors. Such bonds will then make actors start to invest in each other and make adaptations of their resources (Dwyer et al., 1987; Ford, 1980; Valtakoski, 2015). The contribution of our outlined model is that it shows that when starting in a turbulent business network, it is possible (even at times likely) that there are adaptations among resources already before a relationship is initiated and that it is these resource interdependencies that will drive the development of a relationship, rather than the opposite pattern.

This elaborated reversed pattern between a business relationship and resources adaptations then implies that we, on occasions, need to reconsider and question the notion of business relationship initiation as a linear, gradual and evolutionary process (i.e. Dwyer et al., 1987; Ford, 1980; Wilson, 1995). Instead, our research is in line with Zafari et al. (2023), as it brings attention to the fact that relationships initiation at times might be more multifaceted and have other characteristics than the ones pointed out in past literature (Aaboen and Aarikka-Stenroos, 2017; Edvardsson et al., 2008; Valtakoski, 2015). Thus, it seems that we need to question if models created to study business relationship initiation (Edvardsson et al., 2008) and development (Ford, 1980) in business networks are adequately equipped for use in circumstances of turbulence in business networks.

The final contribution of this study is also that it, in fact, highlights different characteristics when it comes to relationship initiation (Aaboen et al., 2017). We can, for example, see in the case that initiation of a relationship does not always involve moving from an unrecognised to a recognised status (Edvardsson et al., 2008). The findings in this study show that an actor can have an expectant status, predicting that another actor will need their resources in the future, thus awaiting the other actor to recognise the anticipated need and move them to an expected status. This finding contributes to research (Aaboen et al., 2017; La Rocca et al., 2019) which suggests that initiation is not a unilateral process (Dwyer et al., 1987) but a process that includes two active actors.

6.3 Suggestions for future research and managerial implications

The model developed in the paper also offers interesting avenues for future research, as well as managerial implications. In terms of future research, we can see at least three possible avenues.

Firstly, it would be interesting to take a quantitative approach investigating how common it actually is that new business relationships are initiated from a starting situation in turbulent business networks, and investigate in quantitative numbers how often business relationships actually start from previous resources adaptations, rather than through the formation of social bonds.

Secondly, given the importance of considering the frequency of turbulent business networks, we believe it is important to further test and adjust models of relationship initiation and development (Edvardsson et al., 2008; Dwyer et al., 1987) to such a starting situation. This would, for example, include a further and more in-depth investigation of the specific aspects identified in this paper (e.g. the role of bounder spanners) and the challenges that actors who try to initiate new relationships in turbulent business networks face. If deciding to take on such endeavour, it can also be of interest to investigate situations with different triggers (Mandják et al., 2015), be it mergers, acquisitions, reshoring or else, to see if the pattern found is the same or different, depending on the triggering situation.

Thirdly, in this paper, we have challenged a “basic” assumption regarding the starting situation for relationship initiation and development (i.e. that the relationships and actor bonds are established before resources adaptations are created). We urge scholars to continue this course of action and, hence, critically investigate and challenge the starting assumptions that underly our contemporary understanding of how a business relationship comes into being.

In terms of managerial implications, the model developed illustrates the importance of actors developing knowledge about the history of the resources in their possession and if they have been adapted to any resources controlled by other actors previously. Resource heterogeneity makes the managerial task of determining how to optimally combine internally controlled resources with other resources available in the network one of the greatest challenges, but also a rich source of business opportunities.

Figures

Figure 1

Model of the starting situation for business relationship initiation in turbulent business networks

Table 1

View in relationship development and initiation literature on the starting situation and when adaptations are argued to appear

Author/model	View of the “starting situation”	Quotes that describe the view of the “starting situation”	When adaptations are argued to arise
Stage models
Ford (1980)/buyer–seller development	Not explicitly mentioned. The model starts in the “pre-relationship stage” with evaluation and reduction of distance	“At this early stage in their relationship, both buyer and seller are likely to have little experience of each other” (p. 343)	Adaptations appear in “the long-term stage”, the fourth of five stages
Dwyer et al. (1987)/ seller–buyer development	Not explicitly mentioned. The model starts in the “awareness stage” with a company starting to pay by “check” (p. 12) and abandons discrete transactions for relational exchange	“minimal investment and interdependence make for simple termination” (p. 16)	Adaptations (transaction-specific investments) appear in the “commitment” stage that is the fourth stage before dissolution (p. 22)
Mandják et al. (2015)/business relationship initiation	The model begins with the “starting situation” in which actors do not know each other, but where there might exist personal/company reputations and personal relationships	“The starting point [situation] for the birth, the process of emergence, is the simple co-existence of the actors in space and time. At this moment actors are not conscious of the existence of one another” (p. 36)	Adaptations are not mentioned as this model only deals with how relationships are “born”. Thus, adaptations are implicitly viewed as developing in later stages of relationships
States
Batonda and Perry (2003)/ inter-firm network relationship development	Not explicitly mentioned. The model begins with the “relationship searching process” in which there are no commitments between the actors	“No commitments at this stage” (p. 1464)	Adaptations appear in the maintenance state, which is the fourth of the five states
Status models
Larson (1992)/development/formation of entrepreneurial network dyads	Not explicitly mentioned. Relationships start in “Phase 1” with preconditions for exchange, in terms of prior personal/firm reputations, and personal relationships but with no ties between companies	“Each [dyad) began as a relatively arm’s length relation, in that no formal transactions had occurred…” (p. 87)	Adaptations between actors start to appear in Phase 2 of the three phases in the model
Edvardsson et al. (2008)/buyer–seller initiation	The model starts with an “unrecognized status” in which the buyer does not recognise the seller	“We define the unrecognized status as the situation when the parties do not know each other, or most importantly, the buyer does not recognize the seller” (p. 343)	Adaptations are not mentioned, as this model only deals with how relationships are initiated. Thus, adaptations are implicitly viewed as developing in later stages of relationships
Polonsky et al. (2010)/ buyer–seller development	Not explicitly mentioned. The model starts with “exploration” (potential relationship/discovery phase) in which the actors initially are unaware of each other	“Firms in the discovery phase lack a common relationship history” (p. 265)	Adaptations start to appear in the “actualization phase” that is the second of three phases in the model
Aaboen and Aarikka-Stenroos (2017) Start-up initiation	Not explicitly mentioned. The model starts with identification of a need and recognition of an opportunity for mutual business	“[Actors’] lack of mutual understanding or deep familiarity with one another, however makes, the initiation fragile… the initiation can easily be terminated” (p. 235)	Adaptations are implicitly viewed as developing in later stages of relationships. However, there must be a “technical fit” between actors for the relationship to be initiated
Other models
Ring and Van de Ven (1994), inter-organisational relationship (IOR) development	Not explicitly mentioned. The model starts with the emergence of a new relationship	“Cooperative IORs emerge when managers bargain over either the production or the transfer of property rights among legally equal and autonomous parties” (p. 92)	Adaptations (transaction-specific investments) emerge as the relationship develops over time
Wilson (1995)/buyer–seller development	Not explicitly mentioned. The model starts in search and selection, with finding a partner being a “critical step” (p. 340)	“The partner is an untested commodity” (p. 340)	Adaptations appear in the third and fourth of five stages in the model
Valtakoski (2015)/buyer–seller initiation	Not explicitly mentioned. The model starts with “initiation”, with the parties having no previous experience of collaboration at all	…“investments, adaptations or commitments cannot influence relationship initiation” (p. 109)	Builds on Wilson (1995) and argues that investments and adaptations emerge later in relationships as they develop

Source:

Inspired by Aaboen and Aarikka-Stenroos (2017), p. 233–234

Table 2

Interviews

Organisations	Saab	Innovatum	Fouriertransform	Saab bankruptcy estate
Types of interview	Semi-structured interview	Semi-structured interview	Semi-structured interview	Semi-structured interview
Number of interviews	13	3	2	2
Titles of interviewees	1. Director of Electrical Control 2. President Product Development 3. VP Product Development 4. Director of Chassis 5. Manager Chassis 6. Director Body 7. Car Line Team Warranty Manager 8. CEO Saab 9. CEO Saab Subsidiary 10. Manager New Business Development (KAM) 11. Director Four-Wheel Drive 12. Manager - Open Vehicle Software 13. Director of Transmission Engineering	14. Business Developer 15. CEO Science park 16. Incubator Manager	17. CEO 18. Investment Director	19. Trustee 1 Saab 20. Trustee 2 Saab

Source:

Authors’ own work