Multipartner alliances among small firms promoted by external managers: Risk and governance mechanisms

Abstract

External managers may be key to setting up and managing multi-partner alliances (MPAs) among small firms, but their role has not yet been integrated in previous literature on risk and governance structures. This longitudinal matched-pair case study investigates the dynamics between relational and performance risks and control-based and trust-based governance mechanisms, in MPAs promoted and managed by external managers. The integration of external managers allows the identification of different levels of trust and control among partners. This is also the case between such partners and the external manager who controls/trusts, who is controlled/trusted, and how this changes over time. Our results show that trust and control are influenced by both the external actor and partner’s risk perceptions; they also complement each other to manage high performance and relational risks in a dynamic process. We extend prior knowledge by showing that the analysis of control/trust complementarity must take into consideration both partner-partner and partner-manager relationships.

Keywords

multi-partner alliances small firms governance mechanisms external manager risk

Introduction

Many small and medium-sized enterprises (SMEs) form multi-partner alliances (MPAs) to access new markets (Agostini et al., 2015; Lavie et al., 2007); most of these alliances are promoted and managed by external actors (Luvison and Cummings, 2017). MPAs are collective, voluntary organisational associations that interactively engage their members in multilateral value chain activities (Lavie et al., 2007), bound by a unifying goal and governed by a single overarching contract (Heidl et al., 2014). As complex relationships among multiple partners, with no direct reciprocity among them, MPAs present particular managerial complications for member firms, especially SMEs (Albers et al., 2015; Holmberg and Lakemond, 2018). In these challenging arrangements, the threat of opportunistic behaviour by partners, and the difficulties in coordinating joint activities are greater than in other alliances (Heidl et al., 2014).

The dynamics among risk, trust and control systems in alliances have been extensively studied within different theoretical frameworks (Barbic et al., 2016; Das and Teng, 2001; Lee et al., 2015; Massaro et al., 2019; Roehrich et al., 2020; Sengün and Wasti, 2009). Transaction cost economics (TCE) theory and organisational theory argue that the risk of opportunistic behaviour, and the risk of not achieving the expected results, can be controlled by ex-ante and ex-post governance mechanisms (Dekker, 2004). Social exchange theory argues that cooperation and exchange patterns could also derive from socialisation, with trust being one of the most important elements (Cook et al., 2013; Mohr and Puck, 2013). Accordingly, empirical studies have considered how, at each alliance stage, partner risk perceptions may influence the choice of governance structure (control- or trust-based) and how they can develop trust, mitigate risk and enhance control (Caglio and Ditillo, 2021; Dekker et al., 2013; Langfield-Smith, 2008). Notwithstanding the considerable progress made in the literature, it remains unclear how risk affects mixed governance structures in situations such as MPAs, where both types of risk are present (Agostini and Nosella, 2019; Holmberg and Lakemond, 2018).

External independent managers are becoming increasingly important for the formation and development of alliances among SMEs. As small companies often lack the resources to build and manage an alliance (Agostini and Nosella, 2019), external independent managers increasingly often serve as promoters, brokers, facilitators or orchestrators. Their roles both in negotiating the agreement and post hoc behaviour have been well studied. Being considered as key players in the success of MPAs, at each stage they seek to manage risks (Luvison and Cummings, 2017), and reduce uncertainty, by connecting partners so that all parties prosper (Kirkels and Duysters, 2010). In spite of this, their role has not yet been integrated into the literature about risk affecting the governance structure of MPAs.

In order to refine our understanding of that governance structure and to respond to the call for a longitudinal view of governance mechanisms and their interplay (Agostini et al., 2015; Roehrich et al., 2020), we integrate external manager roles into the study of the dynamics among risks, control and trust in MPAs. We address two research aims: first, we investigate how partner risk perceptions influence the design/evolution of MPA governance structures; and second, we examine how the external promoter/manager affects the influence of partner risk perceptions and the design/evolution of these governance structures.

We conduct a longitudinal (Hassett and Paavilainen-Mäntymäki, 2013), matched-pair (Piekkari et al., 2009) case study of two exporting MPAs promoted and managed by external actors, selected for their usefulness in elucidating the dynamics of MPAs (Halinen and Törnroos, 2005). These MPAs work in Spain, where more than 99.8% of the companies are SMEs, and various government programmes (national, regional and local) have encouraged the creation of export alliances. Both are located in Andalusia, but they differ in scope, number of partners and partner’s previous knowledge about each other and the external promoter.

Our focus on MPAs promoted and managed by an external actor offers a new point of view on the complex association among risks, control and trust. First, it has been assumed that partners base their alliance structure on assessments of the risks, the potential advantages of control and the level of trust in the potential partners. We add that trust in the external manager/promoter can also drive this choice. This consideration expands on previous studies which, in accordance with agency and transaction costs theories, link the gradual relaxation or enforcement of controls over the life of the alliance with the increase or decrease of trust among partners (Long and Sitkin, 2018; Roehrich et al., 2020; Sengün and Wasti, 2007). Our results, however, show that the evolution of controls can also be motivated by trust, or the lack of it, in the external manager. Second, our evidence highlights that the dynamic interplay of risk management and governance mechanisms must be analysed at multiple levels to transcend the single-level ‘blind spot’ identified by Lumineau and Oliveira (2018: 23). As MPAs involve a large number of interfaces among partners, and between partners and the external agent, they generate multiple flows of trust and control to be understood. We focus on the flows of control and trust in two different relationships, partner-partner and partners-manager, and our results reveal how the dynamics at one level of analysis have implications at the other level, with both affecting risk management. Third, because transaction costs and organisational theories consider different kinds of risk in the choice of governance form (Billitteri et al., 2013), most previous research examines the influence of each risk in isolation (Caglio and Ditillo, 2021). We illustrate that in each MPA phase, relational and performance risks could share origins and jointly affect their governance, so analysing them separately limits understanding.

The next section details the main theoretical background of this study. The third section develops the empirical study and then discusses the results. The last section emphasises our main conclusions, notes the study’s limitations and suggests future research avenues.

Theoretical background

Risks affecting governance structure

Authors including Das and Teng (2001) and Nooteboom (2004) argue that control systems and trust work together to decrease the risk of failure in strategic alliances. Das and Teng (2001) differentiate two elements of perceived risk: the relational risk that partners will deceive each other, and the performance risk that the alliance will not deliver the expected business results. Studies have predicted how partner perceptions of risk can influence governance mechanisms (Lee et al., 2015), and some of them have suggested that these concerns and governance choices vary over the alliance phases (Barbic et al., 2016; Shen et al., 2020). In fact, partner efforts to manage risk extend far beyond the initial decisions made at an alliance’s formation and therefore, governance structures are reconfigured throughout its lifecycle (Sutton and Brown, 2021). From an organisational theory perspective, Dekker (2004), distinguishing ex-ante and ex-post control mechanisms, claims that control systems are used to motivate alliance members to achieve desirable or predetermined outcomes, and also to coordinate – that is, to clarify mutual expectations, enable goal congruence and establish common ground. Ex-ante mechanisms, such as partner selection or contract design, mitigate risk perception by aligning partner interests, and by reducing needs for coordination during the formalisation stage (Dekker, 2004). Billitteri et al. (2013) found that high relational risk leads towards more hierarchical forms of governance. Control tools manage not only relational risks from opportunistic behaviour but also performance risks from lack of coordination or inability to adapt across company boundaries (Ding et al., 2013). As ex-ante formal mechanisms are often incomplete, ex-post mechanisms may be added to monitor and coordinate partner actions and results once the alliance is up and running (Sanchez et al., 2014) after the formalisation stage.

Searching for, evaluating and selecting a partner comes first (Dekker et al., 2013; Sanchez et al., 2014). In the formalisation stage, these processes are critical (Reusen and Stouthuysen, 2020; Todeva and Knoke, 2005). Even superior management may not outweigh poor partner screening (Cummings and Holmberg, 2012), so information gathering is very important (Reusen and Stouthuysen, 2020) with selection criteria being categorised by Das and He (2006) as task- and partner-related.

Formal contracts specify the detailed rights and responsibilities of each party, the outcomes to be delivered, procedures for monitoring and penalties for non-compliance, and conflict resolution procedures (Camén et al., 2011; Shen et al., 2020). The general understanding is that exchange hazards can be controlled by drawing more complex contracts (Nooteboom et al., 1997). Ding and colleagues (2013) posit that a key reason why contracts differ in complexity is the need to mitigate and monitor varying risks. However, several authors (Blomqvist et al., 2008; Dekker, 2004) claim that, even if the partners clearly understand their relationship’s objective and mutual interests, it is not feasible to design a contract that anticipates all possible eventualities. Some (Dekker and Van den Abbeele, 2010; Sanchez et al., 2014) even argue that partner selection is a learning process that enables a firm to design other enhanced control systems. After formalisation, ex-post control systems (policies and procedures used to ensure that the partner’s behaviour and decisions are consistent with objectives and strategies) complement ex-ante mechanisms and have a twofold function: to control the risks of opportunistic behaviour; and to coordinate activities and resources across companies (Velez et al., 2008).

As controlling for every detail and contingency that may arise within the alliance seems nearly impossible, partners are forced to rely on trust (Blomqvist et al., 2008; Sengün and Wasti, 2009). This view derives from social exchange theory and emphasises the role of informal elements in governance (Cook et al., 2013; De Man and Roijakkers, 2009). Trust can be defined as a generalised expectancy held by an individual that the promise or statement of another individual or group, oral or written, can be relied upon. Trust, in the good will of others, and competence are prerequisites of alliance success (Blomqvist et al., 2008; Mohr and Puck, 2013). In self-organised alliances, partners share their vision and objectives, motivating them to collaborate and enabling them to adapt to each other. Trust prevents opportunistic behaviour, reducing bargaining costs and conflicts; facilitating open communication, mutual understanding, and cooperation; driving the commitment of more valuable resources; and stimulating information sharing between partners (Holmberg and Lakemond, 2018) and learning (Massaro et al., 2019).

The literature suggests that where performance risk is higher, trust-based governance will guarantee the flexibility needed to face it (De Man and Roijakkers, 2009; Dekker et al., 2013; Sengün and Wasti, 2009), and where relational risk is higher, control-based governance will safeguard against opportunistic behaviour (Billitteri et al., 2013; Dekker et al., 2013). However, there is no consensus on what happens when both risk types are present (Long and Sitkin, 2018). Billitteri et al. (2013) find that when performance risk is high, managers prefer more flexible governance solutions, no matter what the level of relational risk, while De Man and Roijakkers (2009) suggest that both kinds of governance mechanisms need to be applied simultaneously. Notwithstanding the considerable progress made in previous literature, theoretical predictions on risk affecting governance structure are not so clear in situations where both types of risk are high, and in which it is necessary to clarify how control and trust are mixed (Long and Sitkin, 2018).

Governance challenges for MPAs

The literature employs different terms to define alliances among multiple partners (Albers et al., 2015), including ‘multi-partner networks’ (Agostini et al., 2015) and ‘multilateral alliances’ (Li et al., 2012). Following Barbic et al. (2016), we use the term MPA to refer to alliances that have a single overarching contractual agreement, shared management and a common objective requiring each partner to interact with the other/s. Thus described, an MPA differs from other collaborations established by SMEs. Involving more than two firms and a dispersed power structure, MPAs are more complex than dyadic (García-Canal, 1996; Li et al., 2012) or hub-driven alliances (Agostini et al., 2015), where power or dependence plays a central role (Nummela, 2003). They also differ from constellations (Gomes-Caseres, 2003), networks (Albers et al., 2015), or alliance networks (Das and Teng, 2001), which involve not only multiple partners but also multiple alliance arrangements, which may have been established separately at varying times and for varied objectives (Li et al., 2012).

MPAs are particularly challenging because both risks are present and need to be managed (Heidl et al., 2014). Compared with other alliances, an MPA carries more relational risk because more partners are independent economic actors, with different, and possibly conflicting, objectives (Gils and Zwart, 2009). This may lead to opportunistic behaviour and cooperation failures (Doz and Hamel, 1998; Groot and Merchant, 2000). When partners are not as crucially dependent on each other as in dyads, it is more difficult to create commitment (Nummela, 2003). They may under invest either out of opportunism or for fear of partner opportunism (Doz and Hamel, 1998; Zeng and Chen, 2003). Moreover, managers perceive even higher risks in alliances composed of competitors (De Man and Roijakkers, 2009) or partners who do not know each other and therefore, cannot anticipate each other’s intentions and actions (Dekker et al., 2013; Langfield-Smith, 2008). Partners may be particularly prone to reject the generalised norm of reciprocity as they may be potential competitors for the knowledge and resources contributed through the alliance, or some partners may have the opportunity to be only takers (Thorgren et al., 2011). Performance risk, both actual and perceived, increases with the number and volume of resources and the diversity of partners, their interdependence and their scope of cooperation (Dekker et al., 2013; Gong et al., 2007; Groot and Merchant, 2000) adding fuzziness, and hindering the attainment of joint targets.

In contrast to other forms of alliances, that allow each dyadic partnership to be managed separately, MPAs are more challenging for SMEs in developing contracts and explicit formal governance mechanisms (Li et al., 2012). Some studies suggest that SMEs tend to create informal alliances, rejecting contracts (Blomqvist et al., 2008; Massaro et al., 2019) and relying instead on trust (Jaouen and Gundolf, 2009). But, as Thorgren et al. (2011) point out, it is difficult to validate trustworthiness for a bundle of firms with different characteristics, intentions and behaviour. Results from previous studies on governance dynamics may not be easily applied to MPAs (Barbic et al., 2016; Lavikka et al., 2015; Li et al., 2012). There are calls for more research to examine the actual practices that small firms use to manage MPAs, how they evolve (Barbic et al., 2016), what influences the use of more control-based or more trust-based mechanisms and how they are intertwined (Roehrich et al., 2020).

The external manager role

A wide range of public programmes support cooperation among SMEs (Nummela, 2003; Virtanen and Hagberg-Andersson, 2017); however, as it lacks a hub and distributes power more evenly among its members, an MPA needs to coordinate its activities and, therefore, needs someone who conceives, designs, leads, develops and manages it, organising the partners (Agostini et al., 2015). Crick and Barr (2007) argue that independent external managers can professionalise this process. In fact, most SMEs seek external consultants to cope with the additional burdens and to overcome their own cognitive and technical limitations (Hanna and Walsh, 2008; Kirkels and Duysters, 2010). The literature on innovation networks has given some attention to the role of intermediaries in this collaboration (whether organisations or individuals) (Hurmelinna-Laukkanen and Nätti, 2018; Kwon et al., 2020). For example, Batterink et al. (2010) found that specialist and independent brokers set up the coordination mechanisms, handle conflicts, enhance transparency, clarify partner expectations, facilitate interactions and are deeply embedded in the firm’s social and business networks. Beyond innovation collaborations, some studies have pointed out that such actors can coordinate marketing (Agostini et al., 2015) or exporting alliances (Virtanen and Hagberg-Andersson, 2017). As SMEs often face difficulties in finding partners, and cooperation with others does not come naturally, these external managers can operate both before and after the alliance is formalised. Before the agreement is established, the external manager may act as a promoter, a strategic sponsor, a visionary defining and promoting the alliance’s dreams (Todeva and Knoke, 2005). The manager helps identify both potential activities and potential members (Virtanen and Hagberg-Andersson, 2017). This person’s ability to find partners with a propensity to cooperate seems to be critical (Huggins, 2000), and Cannatelli and Antoldi (2012) show that the manager helps build trust among partners.

Once the MPA has been formalised, in the operation phase (see Agostini et al., 2015), the literature emphasises the importance of a coordinator. Having a large number of partners reduces the opportunities for them to interact, potentially decreasing their commitment (Li et al., 2012; Nummela, 2003) and increasing their opportunistic behaviour (Fonti et al., 2017). When MPAs are formed upon the initiative of an external actor, there is a risk that partners will favour their own interests over the alliance’s objectives (Wincent et al., 2013). Thus, in the post-formalisation stage, a leading actor is needed to guide alliances to a successful end (Müller-Seitz, 2012). By the same token, a manager’s incompetence could jeopardise the functioning of the alliance (Virtanen and Hagberg-Andersson, 2017). Its development may be strongly linked to the perceptions of the external actor who promotes and/or by manages its initial stages (Huggins, 2000; Lopez-Navarro et al., 2013). This person has a unique and holistic perspective on the cooperation process (Hanna and Walsh, 2008) and can therefore, influence the partner’s choice of governance structures.

Scholars have called for particular inquiry into how external managers develop and manage MPAs despite their lack of authority (Agostini et al., 2015; Holmberg and Lakemond, 2018; Virtanen and Hagberg-Andersson, 2017). Heidenreich et al. (2016) and Landsperger et al. (2012) have studied their main skills, which are related to communication, coordination, the exercise of authority and control and how to moderate divergent interests. Luvison and Cummings (2017) claim they manage the various risks inherent in alliances; their role varies, depending on the degree of experience amongst the participating firms. Inexperienced partners need more support and are much more dependent on the manager’s competence (Virtanen and Hagberg-Andersson, 2017). Even though it has been observed that MPA partners may rely on the actions of an external manager to address risks that arise during the venture (Luvison and Cummings, 2017), and that they rely on their ability to reduce partner uncertainties (Kirkels and Duysters, 2010), their role has not been incorporated into previous studies on risk, control and trust dynamics in SMEs alliances.

Theoretical summary

The debate on the dynamics between control, trust and risk is not resolved in a context of MPAs characterised by high relational and performance risk. The risk implicit in complex relationships among multiple partners with no direct reciprocity makes it difficult to predict theoretically how risk affects MPA governance structures. Most MPAs among SMEs are supported and managed by external entities or managers. This actor’s decision-making and activities are essential in achieving the objectives of the alliance and the interaction among partners. Partner perceptions of the manager’s competencies and goodwill may influence not only their decision whether to participate, but also their perception of risk during the development of the alliance, and even the definition of the governance structure. It is therefore, also relevant to focus upon the external manager to analyse the dynamics between relational and performance risks, and between them and control-based and trust-based governance mechanisms during and after its formalisation.

We have developed a multi-theoretical approach to study how the external promoter/manager and s/his relationship with partners influence on risk perceptions, the design of control- and trust-based governance structures, and their join evolution. Considering that the control-trust dynamic can be present in any relationship where one exchange actor relies on, and is seeking to influence, the activities of another exchange actor (Long and Sitkin, 2018), we combine track-bound (gap-spotting based on existing literature) and disruptive (multilevel analysis as a novel research perspective on MPAs) modes to formulate our research propose (Sandberg and Alvesson, 2011). Therefore, we establish the following research questions focused on both formalisation and post-formalisation stages:

RQ 1

How do partner risk perceptions influence their design/evolution of MPA governance structures?

RQ 2

How does the external promoter/manager affect the influence of partner risk perceptions and the design/evolution of MPA governance structures?

Empirical study

Longitudinal matched-pair case study

Our aim is to add to the existing literature on MPAs, by focussing our study on both the complex and dynamic association between risks, control and trust, and the external manager role. We carry out a longitudinal matched-pair case study (what Buck and Shahrim [2005] call a dual-track case study) on two Spanish exporting MPAs, composed of small independent companies and promoted by external agents. Both were created during the financial crisis (2008–2015), when low domestic consumption pushed Spanish firms to accelerate exports and seek niche international markets.

Although robust findings exist in prior research on SMEs cooperation, we argue this empirical method fits our research aims to gain an in-depth and contextual understanding, by refining current findings, filling ‘empty spaces’ in theory (Pratt, 2008: 248) or investigating ‘a different question’ (Edmonson and McManus, 2007: 1169). In accordance with Welch et al. (2011), we propose it mainly as a ‘natural experiment’. Our study is intended to be explanatory (Yin, 2014), examining causal mechanisms, processes, and links between elements; we combine positivism with a moderate form of the social construction of reality, based on the contextual understanding of managers’ subjective experiences (particularly, perceptions of risk and control, and feelings of trust). In order to support this causal analysis and to increase the external and internal validity of our research, we combine real-time and retrospective methods (Leonard-Barton, 1990), and fit the study’s time frame based on theoretical reasons according to our research questions (Hassett and Paavilainen-Mäntymäki, 2013), beginning shortly after each MPA was formalised and ending when the MPA was sufficiently consolidated.

We chose two recently founded MPAs, both in the agri-food sector in Andalusia, where the impact of the 2008–2015 crisis was greater than in the rest of Spain; with the highest unemployment among the 28 EU countries, in 2015 the region accounted for 18% of Spain’s population but 25% of its unemployment. Operating within the same sector, it allowed us to compare findings for this matched-pair case study. One of the associations focuses on the design and development of olive oil production facilities, the other on the production and commercialisation of gourmet products. Overall, during this economic crisis, the agri-food sector performed well, maintaining its production levels, thanks to the fact that it compensated for the decreases in domestic demand with the growth of its exports.

We began by collecting information about the population of MPAs in the sector through a government agency that supports internationalisation. To select our pair of MPAs, we then applied the following criteria to match them (Leonard-Barton, 1990; Piekkari et al., 2009). Both should have been founded between four and six years before data collection; power should be distributed among partners rather than concentrated at a hub; and each should be promoted/managed by an external manager. Each firm should, however, differ in the number and heterogeneity of partners and their scope, and so, in degrees of risk whilst they should also differ in the extent of the partner’s previous knowledge about the promoter/manager. These criteria respect the theoretical sampling approach suggested by Eisenhardt and Graebner (2007) and allow us to contrast the results of the cases (Yin, 2014).

Case 1, established in 2013, is a MPA formed by 19 suppliers of equipment and services within the olive oil sector. This MPA was founded to develop international projects offering business solutions in the different phases that make up the olive value chain. Its activities cover all the activities of this chain (from olive tree planting and the creation of olive mills, to the commercialisation and positioning of virgin olive oil in the markets and the management of by-products). The different activities are undertaken by firms belonging to very different areas, such as engineering, a plant nursery, electrical business, a glass business, an iron company and a consulting firm, among others. Fifteen were located in different provinces of Andalusia, and four in other regions; their main market targets are China, north Africa and India. The promoter is a professional, previously unknown to the partners.

Case 2 was formed in 2012 by five companies from the agri-food sector (wines, cakes, cheeses, fish preserves, ham and olive oil), which produce or manufacture their gourmet products in a handcrafted way. Their main objective was to expand sales of their complementary products into foreign markets, such as Germany, France and Italy. All are small family firms, located in villages of the same province of Andalusia. The promoter was the manager of a foreign trade institution, considered a reputable professional, and previously known to the partners.

Data collection and analysis

We collected qualitative data from non-participant direct observations, informal conversations, interviews and archival evidence, methods well suited for within-case and across-case analyses (Ridder, 2017) and for real-time and retrospective cases (Leonard-Barton, 1990). Documents include reports on partner meetings; slides of from such meetings; contracts; sales reports; and information extracted from web pages, newspapers and emails. They date from both the formalisation stage (in which the external manager plays the role of promoter) and the first years of operation (in which the external actor plays the manager role). From 2012 to 2015, we conducted 21 interviews (Table 1), each lasting between two and four hours, with three export consultants (focused on collecting background data about both MPAs and their members), eight small-firm owner-managers and the two alliance managers (two separate rounds of interviews). Interviews were semi-structured and open. Researchers analysed the qualitative data in an ongoing, iterative and cyclic process to adjust the initial interview protocol, adding new issues according to unexpected responses and emergent themes (Edmonson and McManus, 2007; Eisenhardt, 1989; Yin, 2014). Two researchers were present at each interview, and all interviews were recorded and transcribed. In addition, researchers took field notes at the interviewees’ facilities.

Table 1.

Information about interviews.

Interviewed	Description	2012	2013	2014	2015
Consultant 1	Public export service agency manager.	4 h
Consultant 2	Freelance consultant, an expert on consortia.		4 h
Consultant 3	Freelance consultant, an expert on consortia.		3 h
Case 1 Promoter/Manager	CEO: Alliance 1 promoter, agricultural engineer, former export manager in another international firm.		3.5 h	2 h	2.5 h
Case 1 Partner 1	CEO: National expert on nursery production of certified and high-quality tree plants.		2.5 h		2 h
Case 1 Partner 2	CEO: International technological partner for the development of advanced projects of agricultural drip irrigation and automation.			3 h
Case 1 Partner 3	CEO: Technological partner for the development of advanced engineering projects.		4 h		3 h
Case 1 Partner 4	CEO: Leader partner in cleaning and washing of tree plants.			2 h	3 h
Case 2 Promoter	Manager of the Chamber of Commerce.	3 h
Case 2 Partner 1	CEO: family business mainly making artisanal cheese.		3 h		2 h
Case 2 Partner 2	CEO: family firm making artisanal bakery products.		3 h	2 h
Case 2 Partner 3	CEO: family business producing gourmet fish products.			3 h	3 h
Case 2 partner 4	CEO: family firm making finely crafted confectionery.				3 h

Given that our data are narrative, complex and polysemous – reflecting the complex and dynamic process we wished to understand (Langley, 1999) – we used a qualitative inductive analysis (Thomas, 2006: 240) to ‘[develop] categories into a model or framework that summarises the raw data and conveys key themes and processes.’ First, we prepared the qualitative data, including detailed and annotated interview transcriptions. All three researchers read the data. Once we had developed a clear sense of the consortia’s backgrounds, contexts and governance structures (see Miles and Huberman, 1984), we analysed our field notes and interview transcripts to identify categories, themes or issues of significance, triangulating them with archival data to improve construct validity (Gibbert et al., 2008; Yin, 2014). Although the conceptual framework and the semi-structured interview scripts were likely to favour the rise of certain themes, the main ones emerged from the initial reading of data. We designed a table to organise and reduce the notes, interview transcripts and archival data into first- and second-order themes for each MPA, and analysed the common issues in the cells, looking for patterns within the data, relating them to theoretical issues. We used this table to revise and consolidate theoretical categories (shown in Figure 1), and devised visual schemes to connect them (Corbin and Strauss, 1990; Langley, 1999; Thomas, 2006). Later, by organising the categories into chronological order, we examined the flow of events to focus on ‘what led to what and when’ (Miles and Huberman, 1984: 194).

Figure 1.

Overview of data structure: first- and second-order themes and aggregate theoretical dimensions.

To achieve internal validity in the patterns of causality between variables and results, we interpreted the qualitative data with reference to the conceptual framework (Yin, 2014), attempting to clarify the chain of evidence when telling the story. We focused both within-case and comparative cross-case analyses on the relations and dynamics addressed by our research questions (Eisenhardt, 1989), emphasising interconnected level of analysis and locating them contextually and temporally (Pettigrew, 1990). To ensure external validity – the potential transferability of this study’s conclusions to other contexts (Cepeda and Martin, 2005) – we have included thick descriptions in the present text, mainly as interview transcript excerpts; following Pratt (2008: 501), we present ‘power quotes’ as quotations in the main text and ‘proof quotes’ in Table 2. Producing a case study protocol has improved the reliability of this research (Yin, 2014), even in its longitudinal character, since data triangulation allows ‘each type of data [to provide] a unique but overlapping perspective’ about ‘what happened as well as when it happened’ (Street and Ward, 2012: 170).

Table 2.

Proof quotes.

Statements about …	Proof quotes	Interviewed
Process in the formalisation stage	My consulting work was to assist partners in guiding them in the constitution process […]. Mentoring is very important to ensure viability. With the aim of starting to run, we used standard articles of incorporation to which we later added more details based on meetings with partners.	Case 2 Consultant 2
Process in the post-formalisation stage	I think we are working well, in a serious way […]. Our market image is now very good, people know us, but it takes time.	Case 1 Partner 4
Idea to establish the MPA	[The idea] came to me working for a company in the olive oil machinery sector in which I was export responsible. So during the development of this activity, demands were coming to me, when you talked to clients [they asked you for] things that were complementary to the offer I was carrying. Seeing that there was such great demand and that the cost for internationalising a company was also high, [when I was unemployed] I thought “Hey, why not?” I could join several companies, share expenses and also be giving customers an offer that they were demanding, that they were asking for something that the more complete the better.	Case 1 Promoter/Manager
Idea to establish the MPA	He was touching several companies, offering us [a chance] to collaborate together, and so the germ was created. At first, he spoke to us about what was intended, about using the public aids that were available, taking advantage of the synergies.	Case 2 Partner 3
Partners selection criteria	I was looking for companies that, well, I didn’t know either, but, because of my knowledge about both the sector and the reputation firms had in the sector, I was looking for companies that […], that worked well or that I knew were well positioned in the market […] The best reference was sought in each process, taking into account also other criteria, such as complementarity with the others, previous experience in exporting (…) I have taken not so much the best ones but the ones who believed in the project the most.	Case 1 Promoter/Manager
Promoter functions	[The promoter] reconciled our interests in our first meeting; he assured us that everyone would benefit.	Case 1 Partner 1
Manager functions	The sales plans are established in the assemblies [with a frequency of 2 or 3 months], by ourselves. The manager proposes to us, “I think we can go to this market, to this fair,” and we approve or not, and he explains to us all the actions and their results.	Case 1 Partner 3
Antecedents of relational risk perceptions	These mechanisms [in statutes] were established to defend the alliance from the bad practices that different companies may have. I think it was all the result of mistrust because most partners never had a relationship or knew each other.	Case 1 Promoter/Manager
Initial level of relational risk perceptions	When [the promoter] was having conversations with us for constituting the group, it seemed a risky venture […] we were a lot of firms with different, shall we say, interests and we didn’t know each other [and] we hadn’t any export experience.	Case 1 Partner 1
Initial level of relational risk perceptions	The synergies between their products helped prevent suspicions about opportunistic behaviours.	Case 2 Consultant 2
Later level of relational risk perceptions (risk evolution)	I think the best thing about our group is that we all get along great, we have a very trusting relationship.	Case 2 Partner 4
Antecedents of performance risk perceptions	As a whole chain […] I work on this part, another firm starts here and finishes there, and everyone complements each other to offer the whole project to the customer, although every partner manages how to do its own operations. If one member fails, the project fails.	Case 1 Partner 2
Initial level of performance risk perceptions	We were all companies from different sectors located in the Sierra of Cádiz, a pretty good group, and we didn’t compete with each other. We had complementary products to go abroad, to go abroad with the consortium […] none of the companies had an export department.	Case 2 Partner 1
Later level of performance risk perceptions (risk evolution)	We are hopeful because, although it is taking time, we are already completing agreements for large projects.	Case 1 Partner 1
	More or less I think we are working well, in a serious way […]. Our market image is now very good, people know us, but it takes time.	Case 1 Partner 4
	We started to think about losses.	Case 2 Partner 1
Antecedents of partner-partner trust	The problem is that each one watched over their own interests, we weren’t interested in unifying efforts and each one took care of its own […] I understand that it was due to lack of trust and ignorance, so it is difficult to make an alliance. We didn’t know each other previously; the promoter brought us together with each other.	Case 2 Partner 3
Types of partner-partner trust	Many partners had never had a relationship with each other, they didn’t even know each other, so trusting each other was difficult for us …	Case 1 Partner 4
Initial level of partner-partner trust	Most of us didn’t have any previous contact […]. For that reason evidently at the beginning there was a little more distance [between the partners], but from having meetings we already knew each other.	Case 1 Partner 3
Later level of partner-partner trust (trust evolution)	I think the best thing about our group is that we all get along great, we have a very trusting relationship we share the essentials in the way we do business […] with a lot of respect.	Case 2 Partner 4
Antecedents of partner–promoter/manager trust	The consortium arose through the Chamber of Commerce. We had a very good contact with whom we had attended international fairs, and today he is a friend.	Case 2 Partner 2
Types of partner–promoter/manager trust	He helped us to integrate our ideas, and resolved conflicts in the contract draft.	Case 1 Partner 2
Types of partner–promoter/manager trust	We credited him with high commitment to the alliance.	Case 1 Partner 4
Initial level of partner–promoter/manager trust	The alliance arose through the Chamber of Commerce, where we had a very good contact with whom we had attended international fairs.	Case 2 Partner 2
Initial level of partner–promoter/manager trust	[The manager] was imposed by the promoter; really, when [the promoter] convinced us about the alliance, we thought he would be our manager, but then he told us that [the manager] would be a good friend of him, with export experience.	Case 2 Partner 1
Later level of partner–promoter/manager trust (trust evolution)	No one has asked me to know in detail how much has been spent, […] I think I have their confidence.	Case 1 Promoter/Manager
	We did not know if he [the manager] were able to get sales, or he was doing his best.	Case 2 Partner 2
	I think the failure […] is in who is hired as manager. This man didn’t know how to translate, well, convert his work into commercial actions. I don’t know the work he did, but he didn’t materialise any sales.	Case 2 Partner 3
Control goals	In the meetings, I explain to them the market situation, the results of our actions and also the upcoming dates of fairs or events […] the planning [and] the summaries of what we are doing. These meetings also serve as a discussion forum between the partners.	Case 1 Promoter/Manager
Control goals	He explains to us all the actions and the results of the actions that are being carried out (Partner 2)	Case 1 Partner 2
Initial control focus	Because we had suspicions we tried to regulate everything.	Case 1 Partner 2
	In our statut.es, there are several clauses that penalise companies that may be using the information to sell outside the group or give our information to third parties and firms that are not working well.	Case 1 Partner 4
	We met periodically at each partner’s premises to establish synergies and our actions plan.	Case 2 Partner 1
Later control focus (control evolution)	In the beginning, I sent them a monthly report …, but over time some partners told me, “Do not send me so many papers […]” I give them the accounts [and] I also inform them by phone or by e-mail.	Case 1 Promoter/Manager
	I don’t have time to examine documents, for that is the manager’s job.	Case 1 Partner 3
	We asked [new manager] for actions and results […] we established more and more reports.	Case 2 Partner 2
	Every X time we met […]. Then we scheduled his attendance at fairs and trade missions, we even tried to pay him [the manager] by means of commissions instead of a fixed salary.	Case 2 Partner 4

Results

Case 1

This is an alliance among a large number of small firms without any previous contact. The promoter had been working as an export manager in another business in the olive sector and he knew of the different firms working in the sector through their good work and reputation. When he lost his job during the financial crisis, he decided to develop his own business idea ‘looking for the best firms in their domain’ (Promoter/manager). He had a deep knowledge of the sector and carried out an analysis of the various Andalusian firms capable of supplying the equipment or the know-how required in each of the required phases of a full turnkey project. He selected firms according to their reputation, reliability and recommendation, as none of their managers knew each other directly. The external promoter sought the best references in each process, complementarity of skills with other partners and some experience in exporting was also valued. Then he selected partners with a better reputation to carry out each activity and convinced the partners to join the MPA. Initially, Andalusian companies were seeking to benefit from the regional government’s aid; as only 75% of the companies were required to be Andalusian, in a few cases the search for partners was extended to the rest of Spain. Most of the businesses initially contacted agreed to participate, all of them having significant experience in the sector.

Partners had high perceptions of both relational and performance risks: ‘The uncertainties about partners’ behaviour were high since companies were sharing the business with several unknown partners that differed in size, experience, structure, etc.’ (Promoter/manager). Some firms had some export experience, and their objectives might clash with those of other partners, or the whole alliance:

When [the promoter] was having conversations with us for constituting the group, it seemed a risky venture […] we were a lot of companies with different, shall we say, interests and we didn’t know each other [and] we hadn’t any export experience, but really we just wanted to try selling abroad. (Partner 1)

Another threat to success was the scope and interdependence of the alliance’s activities, which covered the full value chain, with each partner being responsible for one part. The MPA offered its foreign clients completely vertically integrated projects, but also each company could offer solutions to individual problems. During the partner selection process, the external manager gathered information about abilities and competences that was shared with other partners: ‘In fact, this process generated a feeling that we could afford our objectives, some level of confidence, between partners’ (Partner 1). This process reinforced partner trust in the manager’s goodwill and competence. Partner 1 recalled that ‘[The promoter] reconciled our interests in our first meeting, he assured us that everyone would benefit’, and Partner 2 said, ‘He helped us to integrate our ideas, and resolved conflicts in the draft contract’.

Initially, trust between partners was reputation-based and very low. In order to mitigate the performance risk perception of partners, the promoter established task-related criteria for selecting firms to execute the assigned activities of the value chain. This mechanism, together with a good reputation for competence, fostered the trust needed among partners: ‘step by step, we overcame our initial suspicions’ (Partner 3). To alleviate perceptions of relational risk, the promoter used partner-related criteria, such as higher long-term commitment to the alliance. Nevertheless, a detailed contract was established to mitigate relational risk perception. This contract exhaustively covered the firm’s operational cost pool, organisational structure, financial shares, problem-solving mechanisms, penalty clauses, order and sales assignment rules, criteria for selecting new partners, decision-making procedures, manager autonomy, target markets and clients, price policy and cost and profit-sharing.

Additionally, some control tools were anticipated and designed for the operational phase, such as periodic meetings and reports on cost and performance. Meetings and negotiations to design the contract helped generate goodwill and trust among partners in this initial phase: ‘Most of us didn’t have any previous contact […]. For that reason, evidently, in the beginning, there was a little more distance [between the partners], but from having meetings we already knew each other’ (Partner 3). Reports created an infrastructure to control the manager’s activities.

In the early years, the promoter continued as an external manager and the MPA’s activities were focused on sales, but most of them were not fruitful. The group did not get any full turnkey projects, although it exported some small projects involving a few of the partners. To carry out these activities, partners made decisions by consensus, negotiating and deciding about proposed actions, and discussing their results.

Partners used formal control systems to hold the external manager accountable for expenses, actions and results. Paradoxically, the high-performance risk that partners had perceived from the beginning caused them to accept these poor results without increasing their perception of risk: ‘we knew it would be difficult, but that is normal’ (Partner 2). Although the partners perceived high relational risk, they did not use the contract to manage their alliance, considering some clauses restrictive and rigid, and regarding them only as guides: ‘most of the companies do not have the contract in mind, and really everything is to be working day-by-day, that’s it’ (Promoter/manager). However, the complexity of the projects required the use of formal management tools, not only to coordinate partner activities but also to define the projects offered, an important use that they had not contemplated in the formalisation stage. In spite of the initial controls, over time, partners demanded less detailed explanations of manager behaviour, dedicating less time to reading reports and even asking not to receive them.

The external manager designed and executed annual plans, acted as a representative with clients, and resolved potential conflicts between partners, motivating partners to adopt a positive view of the alliance: ‘He continuously encouraged us to work together’ (Partner 1); ‘His enthusiasm was contagious’ (Partner 4). He continued fostering dyadic relationships with each partner, acting as a mediator between them. Frequent interaction with the manager and his ongoing explanations increased partners’ optimism about the alliance’s future, and their trust in his competence and goodwill. Despite poor results, partners’ perceived performance risk did not increase, and the manager’s safeguarding role lowered perceived relational risk.

Case 2

During the economic crisis, a manager of the local Chamber of Commerce saw a potential opportunity for small companies with little prior inclination to work together, and told each one how they could collaborate, motivating them to do so. The alliance had five partners offering homogenous and complementary products, but without any international experience. Their domestic gourmet market experienced a drop in sales of between 4% and 20% during the period 2007–2011. The MPA searched for synergies to reduce and share costs, such as sharing an export department and using a new brand within a distribution channel. These partners did not have previous relationships, but they knew each other because of geographical proximity, entrepreneurial association and Chamber of Commerce activities: ‘I did know the brands or the companies, [but] I had not worked with them’ (Partner 4); consequently, ‘we were not very worried, but we had certain concerns about our partnership’ (Partner 3). Their lack of export experience and similarity in size and culture (all partners were small family businesses) prevented perceptions of conflicts of interest. Taking other successful alliances that the promoter had shaped as a reference, partners trusted the promoter’s export abilities, overcoming their lack of knowledge to establish the alliance.

During the negotiations, the promoter hired a consultant to design the governance mechanisms, enhancing partners’ feeling of ‘being in safe hands’ (Partner 3). The promoter selected partners from companies pertaining to the same local area, searching for product complementarity in order to prevent opportunistic behaviour. Following the consultant’s established procedure, which involved a meeting among partners, the partners signed a contract with standard clauses that regulated only their initial financial investment and their decision-making procedure. The consultant’s protocol was designed to mitigate initial reticence and foster goodwill-based trust among partners. In this way, initial trust was generated through continuous interaction and negotiation. Meetings and reports were planned for the following stage. The consultant produced a programme to jointly develop control mechanisms to govern the alliance.

Partner 2 reports, ‘We shared a high degree of trust in the promoter’s abilities and interests, accepting all his recommendations and suggestions’. Because ‘he was the professional’ (Partner 1), the consultant’s and promoter’s criteria prevailed in each decision in this process. Although some initial trust was based on local reputation and belonging to the same associations, low goodwill-based trust among partners dominated this case.

Once the MPA was constituted, an independent manager, recruited and hired by the former promoter, led the alliance during its first post-formalisation years. This new manager designed and executed annual plans in an independent manner, ‘without taking into consideration our objectives’ (Partner 1). Partners participated in decision-making through meetings where they negotiated and established actions, discussing their results, and where the manager gave formal explanations about costs, activities and results. But partners did not secure any sales through the alliance, and the poor results, added to their less formalised contract, meant that they had to assume unplanned expenses and unexpected commitments, increasing their perceived performance risk. Poor results provoked some doubts, not only about the manager’s competence, but also about his intentions and commitment. ‘Our meetings with the manager became very tense’ (Partner 2). Partners doubted whether he knew their business well enough to commercialise their products, and they also began to be suspicious of him. As a consequence, partners increased their use of formal control systems to monitor their external manager and introduced output control systems.

When poor results continued, the partners dismissed the manager and hired another. They implemented tighter control systems to monitor the new manager’s activities. The persistence of bad results, even after the manager was replaced, increased partners’ performance risk perceptions. However, problems and the search for solutions brought the partners closer, generating a sense of belonging, joint values and mutual competence-based and goodwill-based trust. Relational risk perception decreased, and the partners avoided alliance dissolution and started to engage in occasional business collaborations, apart from the alliance: ‘The relationship between companies is great. In fact, I have gained something, now we are collaborators. [A partner] has a large store where he now also sells my products, when he used to sell the products of my competitors’ (Partner 1).

Discussion

Our two cases differ in initial transactional and partner characteristics as sources of risk and also in initial familiarity between the external promoter and partners (see Table 3). In both cases, the number of small firms involved induced initial perceptions of relational risk. In Case 2, this perception was attenuated, because strong compatibility leads export partners to expect better results (Nummela and Pukkinen, 2007), and in this case partners knew each other from being located in the same province (Dekker et al., 2013; Langfield-Smith, 2008) and considered themselves to have similar cultures (Delerue and Simon, 2009). In Case 1, however, perceptions of relational risk intensified, because some partners had previous experience in internationalisation while others did not; partners perceived this imbalance as a potential source of conflicting interests. That is, whereas limited export competence might be expected to generate performance risk (Billitteri et al., 2013), in Case 1 it also generated relational risk. Supposedly, performance risk and relational risk are independent, having different sources (Das and Teng, 2001), but in our cases the risk sources overlap.

Table 3.

Cross-case analysis in the formalisation stage.

	Case 1	Case 2
Initial sources of risk
Transaction characteristics
Heterogeneity and complexity	19 partners, product and service suppliers including all the value chain links belonging to different sectors.	Five partners, gourmet food manufacturers.
Heterogeneity and complexity		Family firms located in the same Spanish province.
Scope	Full turnkey projects.	Adding products to reach foreign markets. Creating an external export department.
Partners’ characteristics
Previous experiences	There were no previous relationships among them.	They knew each other but there were no previous business relationships among them.
Previous experiences	Some firms had previously exported their products.	No export experience.
Initial trust in each partner	Scant trust based on reputation.	Weak initial trust, formed by reputation and sparse previous relationships.
Partners’ perception of risk
Relational	High.	Medium.
Performance	High.	Medium.
Governance mechanisms
Control-based
Partner selection criteria	Reputed competence and specialisation within the value chain.	Geographical proximity and previous professional relationship with the promoter.
Partner selection criteria	Experience exporting and commitment to the alliance.	Complementary products that could use the same export channel.
Contract	Exhaustive and detailed.	They left many issues open.
Contract	Starting from a standard contract, the partners supplemented it with numerous detailed agreements and penalty clauses.	Standard contract prepared by a consulting firm. Partners followed the promoter’s and consultant’s suggestions to complete the contract.
Formal control systems	Quarterly meetings were planned.	Monthly meetings were planned.
Formal control systems	Detailed monthly reports on cost and performance were anticipated for monitoring.	General reports on expenses were anticipated for monitoring.
Trust-based	Trust grew slightly through contacts and meetings during negotiations and the reputation created by the manager.	Trust grew through contacts, meetings and negotiations, overcoming some initial reticence.
External actor (promoter)
Profile	Technical agronomist, with experience in the sector.	External promoter with experience as an export consultant.
Previous relation with partners	No initial relationship with any partner.	All partners had previous relationship with the promoter and a strong initial trust in him.
Role	Defined the scope of the alliance and selected the number of participants.	He defined the scope of the alliance and selected the number of participants.
	Pivotal. Collected information about potential partners, selecting and convincing the partners.	He convinced the partners, defining their brand and leading the constitution process.
	Mediator in contracting process.	Mediator in contracting process, hiring a consulting firm that acted as an enabler.
Partner trust in promoter	Trust grew, because partners recognised the promoter’s competences, business idea, and commitment.	Strong trust grew because partners recognised the promoter’s competences and goodwill.
Partner control over promoter	Detailed monthly reports and meeting were anticipated.	No controls were established.

In each case, external promoters had a relevant influence on the partner’s initial risk perceptions. The promoters defined the scope of the alliance and selected the number of participants (large in Case 1, small in Case 2) and their diversity (a heterogeneous group in Case 1, a homogeneous group in Case 2). As Dekker et al. (2013) and Langfield-Smith (2008) argued about previous inter-partners collaborations, a history of collaboration between partners and promoter also influences the initial perception of the promoter’s trustworthiness and their assessment of partnership risk.

Dynamics during formalisation

Table 3 summarises the main similarities and differences between cases during the formalisation stage in risks, governance mechanisms, the function of the promoter and their relationship with the partners. During this stage, control-based mechanisms, such as contracts and partner selection, created the initial conditions for cooperation. As the partners had little (Case 2) or no (Case 1) previous knowledge of each other, the promoters collected information on them. In both cases, the promoters were influential conduits of trustworthiness because of their ability to share trust-relevant information. Using formal selection criteria to negotiate affected partner perceptions of each other’s suitability and competence; this persuaded them to collaborate. Once partners were selected, promoters intentionally used the contracting process to boost the necessary trust among them. Initial trust between partners was not evident (Case 1) or very weak (Case 2), but was enhanced by the formalisation process. As Heidl et al. (2014) suggest, the contracting process allowed partners to gain a better understanding of each partner’s roles and responsibilities and improved group cohesion. In accord with the organisational learning perspective (Dekker and Van den Abbeele, 2010), information acquired during the partner search (Reusen and Stouthuysen, 2020) and contract design increased competence-based trust among partners. Repeated interaction and the negotiation of the contract helped partners to align their goals and to build goodwill-based trust. So, extending the findings of Cannatelli and Antoldi (2012), this study demonstrates that, when trust is low or nil, the external promoter intentionally builds it by using control mechanisms.

Although Li et al. (2012) argue that it is more difficult to develop an explicit MPA contract when there are many alliance members, in our cases the larger MPA negotiated and wrote the most complex and complete contract, with detailed controls, while the smaller designed the simplest contract (see Table 3). Generally, in Case 1, the high perceived relational risk motivated the anticipation of additional control mechanisms to achieve and maintain congruence among the diverse goals of partners, restrain opportunism and ensure reciprocity. Confirming the transaction cost theory prediction (Ding et al., 2013), contract complexity and anticipation seem to increase with partners’ initial perception of relational risk.

Roehrich and Lewis (2014) predict that performance risk, derived from the complexity of the MPAs, will lead to significant use of trust-based governance structures. However, Case 1 shows high performance risk leading to more formal partner selection criteria, focused on competence, and to more formal control mechanisms to anticipate coordination challenges even when complexity makes it difficult to specify coordination requirements. As Gulati et al. (2012) suggest, and contrary to the findings of Massaro et al. (2019), in both cases, but more in Case 1, partners included clauses, and reinforced the contract with other formal control mechanisms – not only to prevent opportunism, but also to prevent problems in dividing activities and executing tasks (Shen et al., 2020). In the formalisation stage, our results (Table 3) seem to confirm the prediction by Billitteri et al. (2013) of the moderation between risks. Where noticeable relational risk existed, partners agreed to use the contract and other control mechanisms to reduce performance risk derived from complexity and the need for coordination. However, our cases also show something else: the influence of the external actor in this process. More specifically that partners’ trust in the external actor influenced their decisions about what governance mechanisms to establish.

The inclusion of independent managers guarantees the professionalism of alliance management and can thus mitigate partners’ risk perceptions (Huggins, 2000). Our evidence reveals that during the formalisation process, partner trust in the external actor’s competence mitigates performance risk. As Table 3 shows, in Case 2, where that actor was considered a trustworthy professional, the partners accepted all his suggestions, and he strongly influenced contract clauses and other control mechanisms. In Case 1, where the initial trust in the promoter was limited, the process was more demanding for the partners, and they became more involved in the control design. In view of the multiple interests to be aligned (Das and Teng, 2001), in both cases the promoter acted as a mediator and facilitator of interaction between the companies. In Case 1, the partners’ weak trust in the promoter made them create control mechanisms to motivate not only other’s behaviour but also to make the behaviour of the future manager more transparent.

We have taken into account that, as Long and Sitkin (2018) argue, control-trust dynamics are present in any relationship in which one exchange actor depends on, and attempts to influence, the activities of another exchange actor. Table 3 reveals interfaces among partners and between the partners and the external manager, with mixed control and trust covering different functions at the two levels: partner-partner and partners-manager. Focussing on who controls/trusts and who is controlled/trusted, Figure 2 summarises the dominant governance flows at each level, during each alliance stage as risk evolves.

Figure 2.

Evolution of performance and relational risks perception, and main control and trust flows.

The formalisation stage diagrams in Figure 2 show that in Case 1, where relational and performance risks were high and partners had limited trust in the promoter, both relationships were based on control – both among partners (by means of very formalised selection criteria, a comprehensive and complex contract, and reports) and from the partners towards the promoter (by comprehensive and frequent reports). In Case 2, where both risks were more moderate, the most relevant element was the partners’ trust in the manager, which influenced the control established among the partners.

Dynamics after formalisation

Table 4 summarises our cross-case analysis for the first years after formalisation. During this period neither MPA achieved its objectives. In Case 1, when the real coordination needs of the joint effort became apparent in the operational stage and the results did not meet expectations, partners nevertheless did not increase their perception of performance risk. This was not so in Case 2, where partners had doubts about the new manager’s capabilities and the effectiveness of established control mechanisms, and therefore, strengthened them.

Table 4.

Cross-case analysis in the first years after formalisation.

	Case 1	Case 2
Alliance outcomes	Poor	Poor
Partners’ perception of risk
Relational	High, although they acknowledge that it has decreased	Low
Performance	Medium	High
Governance mechanisms
Control-based
Contract	The contract is only a legal framework, with little operational use	The contract is only a legal framework
Contract	Some clauses are considered restrictive and are dormant. It is not used to govern the relationship	The contract is only a legal framework
Formal control systems	Quarterly meetings	Monthly meetings
	Formal planning	Formal planning. Joint commercial strategy
	Operational planning tools are used to coordinate partners’ interrelated activities	Partners plan actions and establish objectives and an incentive system for the manager
	Project manager tools are introduced to define projects offered and coordinate them	Reports on expenses and performance become more frequent and detailed
	Reports on cost and performance become less frequent; the content is less detailed each time	New reports are introduced
	Reports are not useful; partners are not interested in them	New reports are introduced
Trust-based	Trust grows slightly through meetings, and alliance planning and control, but is not high	Trust grows strongly through contacts and prevents their dissolving
External actor (manager)
Profile	The alliance manager is the promoter	There are two managers during the period studied, neither of them the promoter
Previous relation with partners	All partners had relationships with the manager in the previous stage	The first manager is recommended by the promoter. No initial relationship with any partner
Previous relation with partners		The second manager is recruited by partners after a selection process. No initial relationship with any partner
Role	The manager does the formal planning	The manager initially does the formal planning, suggesting actions
	He leads the projects, acts as a representative	His main activity is sales; he searches for new opportunities and represents the alliance with clients
	Mediator among partners; he coordinates the actions of each partner and motivates them
	Use project management tool
Partner trust in manager	Trust grows strong, despite poor results, because partners recognise his competence and commitment with the alliances	Trust is broken. Partners feel fooled
Partner control over manager	Reports become less frequent; the content is less detailed each time	Reports showing his results and expenses become more frequent; the content is more detailed each time
	Reports are not useful	New reports are introduced
	Partners are not interested in them	New reports are introduced

In both cases, however, relational risk perceptions decreased, as control mechanisms made participants’ behaviour visible and goodwill trust, built during the formalisation stage, improved. In Case 1, because the partners interacted infrequently (see also Zeng and Chen, 2003), perceptions were dictated by the behaviour of the majority of partners. Although the partners initially considered the contract as a safeguard, they did not use it as a governance mechanism, and replaced it with reports or monitoring systems. One explanation might be that, as Barbic et al. (2016) suggest, in the formalisation stage negotiating and writing the contract helped clarify ideas and build trust, but – contrary to Shen et al.’s (2020) claim – at that stage they did not know enough to make the formal contract really useful later on.

Perceived performance risk appears to evolve in tandem with partners’ trust in the external manager. Trust evolves as partners’ observations increase – or decrease – their confidence that the external manager is working actively to achieve the alliance objectives. Table 4 shows significant differences with respect to this evolution. In Case 1, partner trust, developed in the previous stage, was maintained and even increased. Partners felt more confident in the manager’s capacity to achieve the alliance’s purposes, and this enhanced their positive expectations about the alliance’s future performance. In contrast, in Case 2, where a new manager was recruited, a lack of trust emerged, increasing the perception that the alliance might fail. To the extent that an alliance satisfies partners’ expectations, they may experience a virtuous cycle of escalating relational commitment and trust (Doz, 1996; Nummela, 2003; Velez et al., 2008). In Case 2, a number of poor results placed considerable stress upon partner motivation and commitment. Yet surprisingly, poor results increased partner involvement to seek solutions and also, their commitment to the alliance. As Nummela (2003) pointed out, in Case 2, mutual trust and social bonds between partners kept them committed to the collaboration.

The external actor also played a key role in the evolution of the alliance’s governance structure after it was formalised. The partners’ trust in the manager enabled him to orchestrate the alliance in spite of his lack of hierarchical authority (see Dhanaraj and Parkhe, 2006). To the extent that partners trusted the manager, he was able to develop and use governance mechanisms. As Table 4 shows, in Case 1, the trusted manager used control mechanisms to coordinate and motivate the partners, and this practice also influenced the perceived utility of the contract. This new perception exposed weaknesses or rigidities in the contract that prevented coordination in a context of high-performance risk. As a result, partners relied instead on the manager to drive them to achieve their objectives. In Case 2, where trust in the new manager was broken, partners established more control tools to monitor him. In this case, although partners felt they did not have the capabilities to manage the alliance, they increasingly assumed responsibility for coordination and decision-making and established more control with this aim.

Barbic et al. (2016) found that most control concerns are addressed in the initial phase of collaboration. The dynamics described above confirm these results only in Case 1, with high relational risk perception and initial lack of trust in the manager (Table 3). Case 2 shows that performance risk perceptions can motivate the addition of control mechanisms even after formalisation (Table 4). These new control mechanisms improved the MPA’s coordination and the effectiveness of its operations, in line with the arguments in Velez et al. (2008). Figure 2 also shows the evolution of risks and the dominant flow after formalisation. In Case 1, the control governance structure was relaxed because the MPA manager’s control complemented it. Partners trusted the manager’s competence and goodwill, while the manager controlled partners to increase commitment and coordinate joint projects. Therefore, the dominant interface was at the partners-manager level. The controls implemented in the previous stage contributed to further built trust in the manager. Increased trust in the manager not only replaced control in the partners’ relationship with the manager, but also reduced performance risk and relaxed controls among the partners, even though trust among them was not increased. Their trust in the manager made it possible to relinquish control to him.

Case 2 clearly shows a different evolution. Performance risk increased as trust in the manager decreased, and the governance structure evolved to one in which there was both a strong trust among partners and strong control over the MPA manager. As Nummela (2003) points out, partners’ expectations of the outcomes of the exchange are related to perceptions of the commitment of others. Greater interaction between partners, and their involvement in dealing with problems decreased the relational risk among them; trust among partners replaced the need for control. The partners took control of the operation and strong controls were established towards the manager. The needed feeling of confidence (Velez et al., 2008) came from these controls over the manager and the prevailing trust between partners became increasingly stronger.

In summary, cross-case analysis (Tables 3 and 4) brings out two levels of interfaces or relationships: partner-partner and partners-external actor. The analysis shows that, in each MPA phase, trust in the external promoter/manager, and specific risk combinations, influenced the mix of control- and trust-based governance mechanisms. This suggests that risk types underlie the choice of governance structure and its evolution, in situations where direct reciprocity no longer exists and there are fewer opportunities to interact (see Li et al., 2012). In this scenario, the manager can help to bridge potential communication and information gaps between partners by using control mechanisms. Furthermore, to the extent that partners trust the manager, they perceive less performance risk, and they rely on the manager to develop and use control mechanisms. These results extend previous studies that, according to agency and transaction costs theories, link the gradual relaxation or enforcement of controls during the life of the alliance with the increase or decrease of trust among partners (Long and Sitkin, 2018; Roehrich et al., 2020; Sengün and Wasti, 2007). Our results show that the evolution of controls can be motivated by trust, or lack of trust, in the external manager. If the manager proves to be trustworthy, formal controls over them and the other partners may be gradually relaxed even if the partners still do not trust each other. In contrast, low trust in the manager will likely encourage partners to rigidly enforce existing controls on them and the alliance, even if trust among partners has grown.

Limitations and suggestions for future research

Our analysis reveals dynamics and effects that can serve as a source of hypotheses and predictions about the influence of the external manager in different MPAs stages to be tested in future research. Extensions of this work could consider each dyadic partner–external manager relationship separately, in order to quantify, in each stage, for example, the moderating effect of the partner´s trust in the external manager: first, the partner’s risk perception generated by the relationship characteristics and second, the effect of risks on governance arrangements or on the governance mechanisms used.

We find that trust in the external promoter or manager act as a substitute for the trust among partners that is needed to constitute and operate an MPA. Future research should go further along this line and ask what risk conditions or contexts make this substitution possible.

We have analysed only Spanish MPAs formed by SMEs and aiming to export. Other studies can analyse MPAs with different purposes, such as R&D, or in other industries or countries. Although the use of external managers is more frequent in MPAs formed by SMEs, large companies also use them, so future research could extend this analysis to MPAs formed by larger firms. In addition, we have considered only the sources of risk argued by transaction costs theory and organisational theory (Dekker et al., 2013); future studies could take into consideration other types of risk or risk frameworks.

We have assumed that within each case, partner levels of perceived risk, as well as trust, are homogeneous among partners. Although our data triangulation suggests this homogeneity, one must be cautious, because the evidence is qualitative and we did not interview all the participants. It would be interesting to analyse cases in which the partners disagree about the external manager´s trustworthiness or about risk evolution. How would such diversity affect the dynamics between control and trust, and the directions of governance flows? Other research has shown how external managers foster trust among partners; it might be interesting to ask how the external manager takes actions and uses control mechanisms to demonstrate s/his trustworthiness in specific contexts. This approach would extend Long (2018) insights on the manager’s use of control systems with this purpose to a setting without hierarchical authority.

Conclusions

Theoretical implications

Theoretically, our findings suggest (1) the strong influence of an external actor promoter or manager on risk evolution over the life of an alliance; (2) the existence of the different partner-partner and partners-manager interfaces, as relationships that are governed by control and/or trust; and (3) the dynamic interplay between them over time. Figure 3 summarises the dynamic relationships in our data, showing both unidirectional and bidirectional influences.

Figure 3.

Dynamic interplay of risk perceptions and governance mechanisms, and the influence of promoter and managers.

Our analysis indicates that at each stage of an MPA the external promoter/manager and partners’ trust in that person, influence their evaluation of risk and the choice of alliance governance, through a complex, dynamic and intertwined process:

1. Formalisation stage: the promoter (box A in Figure 3) influences partner risk perception when s/he (1) determines transaction and partner characteristics for the MPA, (2) selects partners, (3) demonstrates trustworthiness to partners and (4) designs a contract and other formal controls, and helps in developing initial trust.

2. Post-formalisation stage: the manager’s role and the current level of partner trust (box C) are, respectively, influenced by (5) trust- and control-based mechanisms and (6) the partners-manager trust developed in the previous stage. In addition, manager participation in operating the MPA and the evolved partner trust, on the one hand, moderate both (7) the relationship between risk antecedents and risk perception and (8) the interaction between risk and governance mechanisms (box D), and, on the other hand, affect (9) the development of partners-manager governance mechanisms.

We suggest that the external actor’s roles and partner trust in that person determine governance structures even more than risk perceptions do. In both stages of an MPA, our cases show how the external actor’s role and the partner’s risk perceptions influence trust and control, which in turn act complementarily to manage high performance and relational risks in a dynamic process over time (Figure 3: boxes B and D). Our evidence shows that formal control-trust dynamics evolve and act, synergistically, on managing risk, presenting different mixes in each stage and situation. Accordingly, analyses of control/trust complementarity must consider both partner-partner and partners-manager relationships, present in an MPA, as Figure 2 suggests. At the partners-manager level, both cases show that, where there is scant trust in the manager, partners impose controls on the manager to complement trust, and to ensure that he works for the alliance objectives. At the partner-partner level, our cases show that, when relational and performance risks are high, high partner trust in the manager can supplement the low trust among partners, so control between partners can be relaxed. Previous studies have emphasised the need for external managers to bring partners into contact, and create trust among partners in order to avoid rigid contracts and thus manage performance risk. Our finds that, when the necessary trust between partners cannot be achieved, control relaxation can still be achieved by making partners trust the manager.

Managerial implications

Since external independent agents are becoming more important for the formation and development of SME alliances, our theoretical findings are also relevant for practice (Corley and Gioia, 2011). Governmental agencies that promote MPAs among SMEs often encourage the latter to hire external managers. Our results suggest that agencies should devote efforts to build partners’ trust in the hired manager, and to establish initial controls to make the manager’s behaviour and results more visible. Agencies must rely on people who are able to build trust, both among the partners and from the partners towards themselves; and the agencies should explain to the SMEs what makes external actors so valuable, so that SME owners/managers can choose a promoter/manager wisely. For their part, apart from their expertise and skills, promoters/managers should seek opportunities to prove that they are trustworthy. Partners who have collaborated with the external agent in the past are more likely to perceive them as trustworthy. External promoters and managers of MPAs should be aware that relational and performance risks are interrelated at each stage of an MPA’s development, and that their own partner selection decisions affect the partners’ risk perceptions. The contract can be difficult to write, and even if it is not effective later on, the process of writing it can build trust and manage risk, at the initial stages. Finally, promoters cannot simply serve as a bridge, and leave the MPA to another manager without ensuring that trust has been established between partners and the new manager. In sum, there is room for a leading actor to manage the control-trust association.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Consejería de Economía, Innovación Ciencia y Empleo de la Junta de Andalucía (P09-SEJ-5061, FEDER-UCA18-107689), Ministerio de Ciencia e Innovación (PID2019-104856GB-I00), Ministerio de Economía y Competitividad (ECO2014-57023-P).

ORCID iD

María L Vélez

Author biographies

Maria L Vélez is Associate Professor of management accounting at University of Cadiz. She has a long-standing interest in trust and control of interfirm relationships, internationalization and entrepreneurship. She has published for both academic and practitioner audiences in a variety of accounting and management journals.

Jose M Sánchez is Associate Professor of entrepreneurship at University of Cadiz. His research focuses specifically on trust and control of interfirm relationships, internationalization and entrepreneurship. His work has been published in a variety of management and accounting journals.

Pedro Araújo is Associate Professor of management control at University of Cadiz. His research is mainly focused on control systems and accountability at contemporary and historical contexts. He has published in journals as Accounting, Auditing & Accountability Journal, International Journal of Physical Distribution & Logistics Management, and Business History.

References

Agostini

Filippini

Nosella

(2015) Management and performance of strategic multipartner SME networks. International Journal of Production Economics 169: 376–390. DOI: 10.1016/j.ijpe.2015.08017

Agostini

Nosella

(2019) Inter-organizational relationships involving SMEs: a bibliographic investigation into the state of the art. Long Range Planning 52: 1–31. DOI: 10.1016/j.lrp.2017.12.003

Albers

Schweiger

Gibb

(2015) Complexity, power and timing in multipartner alliances: an integrative review and research agenda. In: Das

(ed) Managing Multipartner Strategic Alliances. Charlotte, NC: Information Age Publishing, 57–87.

Barbic

Hidalgo

Cagliano

(2016) Governance dynamics in multi-partner R&D alliances: integrating the control and coordination perspectives. Baltic Journal of Management 11(4): 405–442. DOI: 10.1108/BJM-09-2015-0173