The role of third actors in the dyadic business relationship initiation: an empirical perspective of sommelier in the wine context

Abstract

Purpose

Starred restaurants, as significant outlets for small wineries, present a unique business opportunity. In this context, the sommelier, as a third actor, assumes a pivotal role in shaping the business relationships between these entities. This study, employing a grounded theory approach, delves into the sommeliers’ roles and activities in the initiation of relationships between small wineries and starred restaurants.

Design/methodology/approach

A qualitative methodology was adopted. Twenty-four semi-structured interviews, direct observations, and informal conversations with starred restaurants, small wineries, and sommeliers were collected and analysed using an abductive approach.

Findings

The findings suggest that the sommelier acts as a contributor to the business relationship initiation between the small winery and the starred restaurant, performing several continuous, simultaneous, and bilateral roles toward both actors.

Originality/value

The study sheds light on the role of wine stewards in the B2B context and provides useful insights to close the theoretical gap between business relationship initiation and the role of third actors.

Citation

Fraboni, P.F.L., Sabatini, A., Marcone, M.R. and Temperini, V. (2024), "The role of third actors in the dyadic business relationship initiation: an empirical perspective of sommelier in the wine context", British Food Journal, Vol. 126 No. 13, pp. 432-454. https://doi.org/10.1108/BFJ-03-2023-0208

1. Introduction

Building a business relationship with a starred restaurant is an intriguing yet arduous business opportunity for small wineries. This channel, highly coveted and crowded, poses significant challenges (Alonso, 2011; Velikova et al., 2019). However, this study proposes that the sommelier (or wine steward) can be a beacon of hope, playing a prominent role as the “third actor” between the two (Aarikka-Stenroos et al., 2018). According to previous academic contributions (Aarikka-Stenroos and Halinen, 2007; Aarikka-Stenroos et al., 2018), the study conceptualises the third actor as a person or an organisation that supports buyer-seller relationship initiation (BSRI).

The third actors’ role in business relationship initiation has been featured in several studies gathered from different industries, such as design, advertising, interpreting, engineering, consulting, accounting, software, banking and shipping (Batonda and Perry, 2003; Mandják et al., 2015; Aarikka-Stenroos et al., 2018; Aarikka-Stenroos and Halinen, 2007). The contribution of Aarikka-Stenroos et al. (2018) and Aarikka-Stenroos and Halinen (2007) suggested that the wine industry is a potential field of future study of the role of third actors. Hence, the study’s originality fills a void by empirically exploring how the sommelier supports the initiation of buyer-seller relationships in the wine industry. Also, by considering the role of the sommelier between small wineries and starred restaurants, the study also highlights the power asymmetry characterising the context (Aaboen and Aarikka-Stenroos, 2017; Siemieniako et al., 2023).

Existing literature depicts the duties and functions of wine stewards (Ruiz-Molina et al., 2010), their influence on wine sales (Manske and Cordua, 2005), and customer satisfaction (Lau et al., 2019), but to the best of our knowledge, no previous studies have investigated their role in initiating business relationships between small wineries and starred restaurants.

By adopting the Grounded Theory (GT) Strauss and Corbin conceptualisation (1990), this article aims to present a novel perspective on how the sommelier supports the small winery-starred restaurant relationship initiation and how the intermediation by a wine steward can foster the interaction between the parties. The study develops a framework based on the technical literature background on the buyer-seller relationship initiation (BSRI) process rooted in the industrial network approach literature, enhancing researchers’ theoretical sensitivity and providing a guide for the data collection and analysis process (Strauss and Corbin, 1990; Dubois and Gadde, 2002).

By observing the role of the sommelier in initiating the business relationship between the small winery and the starred restaurant, the study contributes to the BSRI field of study by answering the following research questions.

RQ1.

How do sommeliers support the small winery-starred restaurant BSRI?

RQ2.

How do the roles of sommeliers unfold in the interaction between the small winery and the starred restaurant in BSRI?

The research adopts a qualitative methodology, entailing semi-structured interviews with three actors: a sommelier, a starred restaurant, and a small winery. Using an abductive approach, the study uncovers the roles the sommelier performs for both actors during the relationship initiation.

The study is organised as follows: section 2 outlines the literature background, section 3 presents the methodology, section 4 describes the findings, and section 5 discusses the theoretical and managerial implications. Finally, in section 6, conclusions are provided.

2. Literature background

2.1 Buyer-seller relationship initiation: conceptualisation

According to the industrial network approach literature, business transactions occur in ongoing, multilateral, inter-organisational buyer-seller relationships involving interactions that create interdependencies, conferring to the business environment a network-like structure (Håkansson and Snehota, 1995; Waluszewski et al., 2019). Business relationships are defined as mutually oriented interactions between two reciprocally committed actors over time (Ford et al., 2011); meanwhile, the BSRI is the process that leads to the first business agreement between the buyer and the seller (Aarikka-Stenroos et al., 2018; Edvardsson et al., 2008; Fraboni, 2023; Sabatini et al., 2021).

The present manuscript focuses on the dyadic BSRI between the starred restaurant and the small winery, investigating the role of the sommelier as a third actor during this process. Despite the numerous studies on buyer-seller relationship development, limited empirical research exists on BSRI, primarily because it is a blurred phase difficult to grasp in real-life contexts (Aarikka-Stenroos et al., 2018; La Rocca et al., 2019; Mandják et al., 2015; Sabatini et al., 2021; Wadell and Bengtson, 2023). Over time, the BSRI literature has produced contributions regarding the business relationship initiation solely considering the perspective of buyers (Claycomb and Frankwick, 2010) or sellers (Edvardsson et al., 2008; Wagner, 2011; Valtakoski, 2015), only limited studies have taken in consideration the dyadic perspective (Ferreira et al., 2017; Klimas et al., 2023). Since every BSRI path is unique and influenced by contextual and environmental factors (e.g. turbulent business landscape, Wadell and Bengtson, 2023; Zafari et al., 2023), the BSRI literature appears fragmented (Valtakoski, 2015): there is no shared consensus about the nomenclature of the phases involved and their progression (for a detailed overview see Table 1). Even though every BSRI path is peculiar and context-specific, it is possible to observe some minor patterns or regularities allowing to systematise the process (Klimas et al., 2023; Sabatini et al., 2021).

Extant research has described BSRI as a progression of stages following sequential growth (Dwyer et al., 1987; Hussain et al., 2020; Klimas et al., 2023). However, limitations have been addressed to the sequential stages approach of the BSRI. It has been criticised for its incremental and irreversible stages progression since the business initiation process can fail at any point (Batonda and Perry, 2003). Indeed, as suggested by Edvardsson et al. (2008), BSRI can start, pause, and end at any point, so the main challenge for business actors is understanding how to proceed further in the initiation process and be considered suitable partners. Therefore, several scholars have developed an alternative perspective on the BSRI process, which is seen as a non-linear, non-sequential, and non-definitive matter of unpredictable states and thus susceptible to early termination (Batonda and Perry, 2003; Ferreira et al., 2017; Hastings et al., 2016; Klimas et al., 2023).

2.2 Buyer-seller relationship initiation phases

Existing studies recognise BSRI development through stages as a viable tool for developing theoretical and managerial contributions (Hussain et al., 2020; Sabatini et al., 2021; Wagner, 2011). Among others, this study will refer to the model of Dwyer et al. (1987). Indeed, as suggested by Hussain et al. (2020, pp. 671), their model is still valid as “many studies follow Dwyer et al.’s (1987) conceptualisation of relationship development and examine the buyer-seller relationship based on lifecycle stages”. The starting point of BSRI is that firms are not acquainted with one another and have a common lack of past shared experience: they only coexist in space and time (Mandják et al., 2015).

According to Valtakoski (2015), BSRI corresponds to the awareness and exploration phases proposed by Dwyer et al. (1987). In the awareness phase, actors are involved in unilateral activities to enhance their positioning and attractiveness to potential counterparts (Claycomb and Frankwick, 2010). Recent literature labelled this phase also as a pre-relationship phase where the relationship does not exist yet; indeed, there are no bonds and no commitment between the actors (Aarikka-Stenroos et al., 2018; Abosag and Lee, 2013; Claycomb and Frankwick, 2010; Klimas et al., 2023; Mandják et al., 2015; Sabatini et al., 2021; Valtakoski, 2015). This phase could persist until trigger events occur (Mandják et al., 2015; Klimas et al., 2023). In literature, these trigger events have been labelled as “converter forces” (Edvardsson et al., 2008), “trigger issues” (Mandják et al., 2015) and “initiation contributors” (Aarikka-Stenroos et al., 2018). After the awareness phase, Dwyer et al. (1987) identified the exploration phase in the BSRI, which is conceptualised in five subprocesses: attraction, communication and bargaining, power and justice, norm development, and expectation development. The five sub-processes enable both parties to test compatibility, integrity, and performance to begin business relationships. Therefore, the exploration phase is characterised by testing and compatibility activities, as the parties are still loosely linked (Ferreira et al., 2017). Reciprocal assessment between counterparts is mainly based on economic and social aspects (Abosag and Lee, 2013; Hastings et al., 2016). During this phase, actors experience uncertainty; therefore, business actors engage in an important communication process to build preliminary organisational and individual trust (Lasrado et al., 2023; Mandják et al., 2015), define mutual goals and relationship boundaries (Hastings et al., 2016; Klimas et al., 2023). This process requires adaptation to the counterpart’s needs, making this phase full of objective and subjective evaluations (Hingley et al., 2008; Mandják et al., 2015; Valtakoski, 2015). The business relationship starts if the parties effectively communicate, negotiate roles that reflect “justice”, and agree on mutual goals and expectations for future interactions (Dwyer et al., 1987).

2.3 Buyer-seller relationship initiation: the third actors’ roles

According to the industrial network literature, the dyadic relationship initiation between the buyer and seller might foresee the role of other actors in the business network. The actors who play a role in supporting BSRI have been termed as third actors of the dyad (Aarikka-Stenroos and Halinen, 2007). In line with Aarikka-Stenroos et al. (2018), the third actor is seen as the “initiator contributor”, supporting relationship initiation between buyer and seller. The third actor can be a firm, a professional (e.g. sommelier), or someone acting in between buyers' and sellers’ interactions (Aarikka-Stenroos et al., 2018).

Previous literature has labelled third actors as relationship promoters (Walter and Gemünden, 2000), actors who are involved in shaping and advancing inter-organisational exchange processes by identifying appropriate partners for different organisations, facilitating dialogue and collaboration between them (Gargiulo and Sosa, 2016), supporting the learning process and solving possible conflicts relying on a rich history of interpersonal interactions with the business parties (Pemartín et al., 2019). Indeed, these actors are acquainted with one or both parties of the emerging dyadic business relationship, and they contribute to the initiation by sharing their experiences and/or by exerting mediation (Batonda and Perry, 2003).

Third actors could minimise the information asymmetry, lack of trust, and risk perception often experienced by partners in emerging business relationships (Aaboen and Aarikka-Stenroos, 2017; Aarikka-Stenroos and Makkonen, 2014; Mandják et al., 2015; Pemartín et al., 2019; Valtakoski, 2015). Aarikka-Stenroos and Halinen (2007) identified twelve roles the third actors can take on during relationship initiation. These are described below.

1. Scouter: suggests potential buyers for sellers and vice versa.

2. Awareness builder: builds awareness for each other actor.

3. Need creator: emphasises the needs of potential buyers.

4. Access provider: offers access to each other party through referral.

5. Accelerator: Accelerates relationship initiation.

6. Advocate seller: supports the seller by attesting to its credibility.

7. Matchmaker: facilitates fit between potential parties.

8. Trust builder: promotes trust by offering an external statement on trustworthiness.

9. Evaluation assistant: helps evaluate the quality of a product/service offered.

10. Expectations builder: helps buyers develop realistic expectations.

11. Risk reducer: reduces risk perception by offering risk-reducing information.

12. Provider of concrete evidence: provides evidence of intangible aspects of the offer.

2.3.1 The third actor: the sommelier

Sommeliers, or wine stewards, are the third actors in the BSRI between wineries and restaurants. The path to becoming a sommelier requires developing expertise in organoleptic analysis and pairing techniques (Festa et al., 2016). Sommeliers play a multifaceted role: they aim to provide customers with the right wine advice, collaborating with wineries while guaranteeing the profitability of restaurants (Manske and Cordua, 2005). According to Ruiz-Molina et al. (2010), wine steward’s tasks can be classified into three blocks. First, creating and updating the wine list and managing the restaurant’s cellar; second, providing a continuous market study; third, fostering wine sales and spreading wine culture.

Sommeliers often act as third actors and independent consultants in small wineries-starred restaurants’ relationship initiation. Their neutrality in judging is granted by the sommelier associations to which they belong. They manage the differences in wine knowledge between restaurants and their guests and between restaurants and wineries (Festa et al., 2016; Manske and Cordua, 2005).

It is important to note that the literature on sommeliers acknowledges other facets. Lau et al. (2019) focus on their role in consumer satisfaction; Dewald (2008) and Ruiz-Molina et al. (2010) on their influence on restaurants’ wine sales; and Dressler and Paunovic (2022) on their capacity to be an important information source for launching specialised wines and for creating hybrid offerings, which is crucial in the present economy.

In a nutshell, the rationale for the study recognises that: (1) in BSRI, there is a need for more empirical studies (Klimas et al., 2023); (2) to the best of our knowledge, empirical contributions to the understanding of the role of third actors in BSRI are scarce (Aarikka-Stenroos et al., 2018), and (3) no empirical studies have highlighted the role of sommeliers in the BSRI between small wineries and starred restaurants yet (Alonso, 2011; Velikova et al., 2019). Therefore, this study develops a field study to unfold theoretical and managerial contributions to the abovementioned literature streams.

2.4 Grounded theory

The study adopts the GT to study phenomena in business and management areas (Birks et al., 2019). The GT is rooted in symbolic interactionism, considering reality as a negotiated phenomenon among individuals always in flux and evolution. The fundamental assumption of the GT is that through a detailed exploration guided by theoretical sensitivity, the researcher can construct a theory grounded in data (Strauss and Corbin, 1990). The primary version of GT was firmly rooted in a positivistic view (Glaser and Strauss, 1967); therefore, it was focused on reducing the possibility of forcing data into preconcerted categories formed by researchers’ previous knowledge and experience. According to Strauss and Corbin’s (1990) versions of GT, the present study recognises that limiting existing knowledge could be counterproductive to the emergence of theoretical sensitivity, which is fundamental for researchers to identify and extract from the data elements which might have relevance for the emerging theory. Since conducting a GT study necessitates an equilibrium between keeping an open mind and the ability to identify elements of theoretical significance (Chun Tie et al., 2019), the study adopts an abductive data analysis process, according to the systematic combining approach (Dubois and Gadde, 2002).

3. Methodology

3.1 Qualitative methodology

This study relies on a qualitative approach, appropriate when the research problem is complex, cannot be easily grasped in a quantitative study, and data must be collected without altering the context (Yin, 2009). The distinctive qualities of qualitative methodologies are flexibility, richness, holism, local meaning, and causality assessment, which are necessary to properly address problematic, dynamic, and multidimensional understudied research topics (Miles and Huberman, 1994). The qualitative methodology was chosen because the study aims to explore a phenomenon in depth and breadth (Voss, 2010) to clarify and extend the understanding of the existing theory in an analytical sense (Kähkönen and Tenkanen, 2010).

Since business contexts are changeable over time, the present study adopts a processual perspective (Langley et al., 2013) that facilitates understanding the sommelier’s roles in BSRI between small wineries and starred restaurants. A field study was developed to collect data from semi-structured interviews, direct observations, and informal conversations. Recognising that no sensemaking strategy is superior to the others, we adopted GT and the “systematic combining” approach - a non-linear path-dependent process to match theory and reality - to analyse data (Dubois and Gadde, 2002) abductively. Integrating GT and abductive reasoning is recommended to yield thorough results in qualitative research. The GT journey entails continuously comparing raw data, codes, conceptual categories, and literature (Charmaz and Thornberg, 2021). The theoretical framework used to guide researchers into the data collection and analysis process has been abductively developed from existing theories on BSRI within the industrial network literature. It served as a heuristic tool to assist researchers in avoiding indiscriminate data collection and enabling them to “dance with data”, thus enhancing their theoretical sensitivity for identifying meaningful data and developing theoretical concepts (see sections 3.3 and 3.4; Chun Tie et al., 2019).

3.2 The research context

The research was performed in Italy (one of the largest wine-producing countries), in the Marche region, where the wine industry has only recently begun to draw attention globally, although its wine-making tradition was traced back to the ancient Greeks. The Marche region’s vineyards cover approximately 15,500 hectares and produce almost 800,000 hectolitres of wine, housing approximately 200 small wineries and a few larger firms. The Marche region entails 5 DOCGs (Denomination of Controlled and Guaranteed Origin), 15 DOCs (Denomination of Controlled Origin), and 1 PGI (Protected Geographical Indication) [1]. Despite its rural setting without significant urban centres or large infrastructure, the region hosted six Michelin-starred restaurants in the 2022 edition and eight in the 2023 and 2024 editions of the Michelin Guide, showcasing its commitment to pursuing excellence in the gastronomy sector. Furthermore, Marche region policymakers have recently launched regional initiatives (e.g. “Dal vino alla Tavola”) that incentivise winemakers to reach the restaurant channel through the mandatory engagement of sommeliers to participate.

As reaching starred restaurants is challenging for small wineries, the Marche region is an interesting context for gaining certain insights (Siggelkow, 2007). Furthermore, the research team’s familiarity with the wine context and geographical proximity resulted in additional rationales for selecting the Marche region since preferential access to data and informants was granted (Yin, 2009).

3.3 Data collection process

The data collection process has been developed simultaneously and iteratively to data analysis to interrogate data and produce early ideas that inform the subsequent data collection conducted through theoretical sampling (Birks et al., 2019). Data were collected primarily through semi-structured interviews (Kvale, 2012). The research team discussed the interview protocol and reviewed it with a sommelier who owns a restaurant to ensure it fits the research aim. The questions were focused on letting the interviewees describe the BSRI unfolding process (Langley et al., 2013). According to the research aim, key informants were selected to be specifically: (1) starred restaurant key person (owners) that have had business relationships with small wineries; (2) professional sommeliers (belonging to official associations and with formal certifications); and (3) small winery owners that have had business relationships with at least one starred restaurant. Key informants have been selected in the business context of the Marche Region as per the previous section 3.2. Informants’ engagement was remarkably difficult, as they are busy and with few available time for the interviews (see Figure 1 for further details).

Finally, the study comprises 24 interviews (see “Data collection process” in Figure 1). Additional data were collected through direct observations, informal conversations, and two events to triangulate the interviewees' information. Data collection was considered exhaustive only when a recurring pattern in the data was found (Dubois and Gadde, 2002) and theoretical saturation was achieved.

3.4 Data analysis

Data were coded manually to answer the research questions. A specific framework was developed (see the picture in the row “Data analysis to answer the second research question” Figure 1) to unfold the roles performed by the wine steward in relationship initiation. The framework allows the depiction of the interaction between the actors: small winery (seller), starred restaurant (buyer) and sommelier (third actor). The abductive examination led to the identification of the sommelier’s roles during initiation (Dubois and Gadde, 2002). The data were compared with the third actor’s roles described by Aarikka-Stenroos and Halinen (2007). The abductive comparison allowed for identifying other roles played by the third actor during the BSRI and new implications for BSRI when the third actors are involved (Figure 1). Below, we provide the sequence of the data analysis activities performed, which are in line with the abductive thematic analysis process described by Thompson (2022).

1. The initial coding activity was descriptive according to the source (sommelier, starred restaurant, or small winery) and the specific subprocess of the BSRI.

2. The second coding activity was open; two researchers coded line-by-line, writing memos about code properties or emerging ideas (Charmaz and Thornberg, 2021). Line-by-line coding was essential to let the researchers deepen participants' experiences and challenge their theoretical preconceptions.

3. The list of generated codes was discussed to identify similarities and differences. The output of this process was selecting a list of focus codes capable of properly representing large batches of data.

4. Cycling forth and back from the theoretical framework and data analysis allows the research team to check the pertinence of the focus code and the related emerging clues with the third actor’s roles described by Aarikka-Stenroos and Halinen (2007) and the phases of the BSRI of Dwyer et al.’s (1987) model.

5. The previous step informed the subsequent data collection and data analysis.

6. Once the new data stopped providing additional insights, the theoretical saturation was considered achieved.

7. The roles emerging from the data analysis compared with the theoretical framework were then analysed to describe in depth what the third actor performs in each phase for each role identified during the BSRI.

8. The theoretical sensitivity guided by the framework developed (see Figure 1) allowed the research team to develop theoretical implications.

9. Finally, to ensure the validity of the results, a draft of the findings was discussed with two wine market experts.

4. Findings

The data emphasise two aspects. Firstly, the BSRI is characterised by blurred phases and an unpredictable path; also, compared to Dwyer et al.'s (1987) framework, the “expectation development” stage precedes the “communication and bargaining” stage. Indeed, after an actor’s preliminary attraction, the development of expectations pushes them to start communicating and validating their expectations. Secondly, during the interviews, sommeliers highlighted their impartial role, which means they are aware that it is a fundamental aspect of their profession.

As sommeliers, we must be distant from favouritisms toward wineries; we must use fair judgment toward everyone; this preserves our professionalism and credibility. Operators ask for services that are as objective as possible.

Sommelier (11)

To nurture the debate, the findings are presented in the phases outlined by the BSRI framework and include relevant quotes from the interviews (see Table 2 for a summary of the findings).

4.1 Awareness – positioning and posturing

Unilateral actions of the actors characterise this subprocess to enhance their visibility within the network without interactions (see Box 1, Table 2). Here, the sommelier detects shortcomings on the starred restaurants’ wine list, helping starred restaurants define their needs.

The sommelier must create a wine list that is responsive to the needs of consumers while also including new products.

Sommelier (5)

Sommeliers highlight shortcomings on the wine lists and suggest new products that could resolve them.

The sommelier has a function of signalling wines to the starred restaurateur. The sommelier does outreach.

Small winery (a)

The sommeliers’ suggestions increase small wineries’ awareness among starred restaurants. Therefore, the sommelier improves the starred restaurant’s ability to identify small wineries.

4.2 Exploration – attraction

Attraction initiates the exploration phase (see Box 2, Table 2). The sommelier helps the starred restaurant evaluate wines that could fulfil its emerging needs.

The wine, although renowned, might not be suitable for the starred cuisine and thus invalidate all the research done to build the wine list.

Sommelier (13)

The sommelier assesses the quality and compatibility of wines with the restaurant’s offering and, if the evaluation is positive, organises a meeting with the winery.

Once a restaurateur told me that he was interested in tasting a wine, I called the winery to organise a visit at the winery.

Sommelier (11)

Similarly, small wineries are interested in meeting starred restaurants as they perceive the opportunity to enhance their products.

I value positively the possibility of dealing with starred restaurants; it gives value to our brand.

Small winery (b)

Hence, sommeliers are also important for small wineries because they support their access to new relationships with starred restaurants.

Sommeliers help us organise initiatives toward starred restaurant aiming to make them taste our wine.

Small winery (d)

4.3 Exploration – expectation development

Expectation development involves the actors in prospecting business opportunities (see Box 3, Table 2). Sommeliers help starred restaurants develop realistic expectations of the wines and their business potential. If the starred restaurant is satisfied, it establishes a business relationship with the winery.

The sommelier creates the mix between wines that are left to age and those ‘ready to drink’.

Sommelier (10)

4.4 Exploration – communication and bargaining

During this phase, the actors show interest in each other’s goals and define mutual obligations (see Box 4, Table 2). Building trust between actors is crucial: the sommelier acknowledged as an industry expert, represents the link for trust building by providing evidence of the small winery’s value proposition.

To accept a new product in the wine list, I personally proceed to know the winemaker or at least the wine.

Sommelier (13)

The sommelier’s information about the small winery reduces the risk perceived by the starred restaurant.

The sommelier has a role of credibility, guaranteeing the quality of the product to the restaurateurs.

Small winery (d)

4.5 Exploration – power and justice

The sommelier manages power asymmetry between actors to ease the emerging business relationship (see box 5, Table 2). The sommelier improves small wineries’ knowledge of starred restaurants’ procurement choices and the deal’s fairness.

A starred restaurant’s sommelier told me that he looks for wines priced at least 40–50 euros hence he does not accept wines whose retail price was far below that figure because consumers – with a quick internet price comparison – can spot an unjustified price disproportion, putting the starred restaurant in trouble.

Small winery (b)

In addition, the sommelier supports evaluating small winery compatibility with the starred restaurant’s business logic.

I hardly buy wines that my sommelier deems not suitable for my business.

Starred restaurant (2)

4.6 Exploration – norm development

Here, the actors define the guidelines for future exchanges through the support of the sommelier. Hence, it allows actors to find a trade-off between their needs to begin the relationship (see box 6, Table 2).

The small winery should be eager to sell small quantities. I know that selling just one box may not be enough, but they should understand that I have storage management costs, so I cannot buy large quantities for all the references on the list […] the small winery to remain on our wine list should comply with restaurant’s business issues.

Starred restaurant (2)

5. Discussion

5.1 Theoretical implications

Study findings suggest that the sommelier playing as a third actor in a dyadic relationship is paramount in facilitating business relationships between small wineries and starred restaurants (Aarikka-Stenroos et al., 2018). During the process, the sommelier plays different roles simultaneously, depending on each specific phase. Furthermore, the study suggests that sommeliers foster relationship initiation with starred restaurants through credibility signalling and link creation (Hingley et al., 2008; Sabatini et al., 2021).

In line with the existing literature on the limitation of the BSRI stage approach, the study shows that the BSRI proceeds through unpredictable states of development (Batonda and Perry, 2003; Hastings et al., 2016; Klimas et al., 2023), where the linear progress of development phases is just one of the possible trajectories of the relationship evolution (Hussain et al., 2020; Klimas et al., 2023). However, relationship evolution as a sequential growth stage demonstrated to be appropriate for representing a complex process that unfolds intertwined and overlaps over time (Hussain et al., 2020; Sabatini et al., 2021). By adopting the framework proposed by Dwyer et al. (1987), the present study finds that the “expectations development” precedes the “communication and bargaining” sub-phase. The present study suggests that in contexts like the wine sector, where product features cannot be evaluated before the purchase, the “expectations development” subphase could be important in leading the parties involved to begin the relationship.

The sommelier could play multiple and simultaneous bilateral roles towards both business parties during the BSRI process (Festa et al., 2016). Concerning the third actor’s roles in relationship initiation described by Aarikka-Stenroos and Halinen (2007), this study suggests that the wine steward performs the following roles: scouter, need creator, access provider, matchmaker, evaluation assistant, expectation builder, risk reducer, and provider of concrete evidence. Moreover, the study identifies a new role that could be assigned to the sommelier – and hence to the third actors – during BSRI processes: the power balancer. This role is important when there is an initial significant power and information asymmetry between actors, which might affect the development of the relationship (Aaboen and Aarikka-Stenroos, 2017; Alonso, 2010; Kähkönen and Tenkanen, 2010; Reynolds et al., 2009; Siemieniako et al., 2023). The sommelier balances these differences bilaterally, allowing both parties to evaluate information flows fairly. Thus, according to the “initiation contributor” notion expressed by Aarikka-Stenroos et al. (2018), the sommelier is the trigger that fosters the initiation of small winery-starred restaurant relationships. Considering the trust builder role, the study suggests that it could be seen as the result of the other roles played by the third actor. The sommelier builds reciprocal and bilateral trust by reducing risk perception and providing concrete evidence (Valtakoski, 2015); hence, the trust-builder role could be interpreted as an effect of the other roles performed by the third actor instead of being a role in itself.

Similarly, the accelerator role could result in the positive development of the BSRI when the third actor is involved rather than a role. Therefore, the study suggests that some roles are preliminary to others (Aarikka-Stenroos et al., 2018). Indeed, some roles emerge at the beginning of the BSRI process (e.g. need creator, scouter, access provider, evaluation assistant, expectations builder, risk reducer). In contrast, others unfold in later stages (e.g. power balancer, trust builder, matchmaker), emerging when preliminary roles have been fulfilled. The role of awareness builder seems to be embedded in the scouter role; starred restaurants usually receive many references for the wine list, so being aware of other wineries does not exert any influence. Instead, the scouting activity incorporates awareness-building activities and indicates what firms would resolve a specific business need. In addition, advocating for the seller can be challenging for third actors (especially sommeliers), as they should maintain neutrality in judging to preserve their professionalism and credibility. “Table 3” (below) summarises the literature contributions mentioned in the study.

5.2 Managerial implications

The study elucidates how the sommelier, as the third actor, supports buyer-seller relationship initiation in the case of starred restaurants and small wineries. The study unfolds a framework to identify the third actors’ roles according to each phase of the BSRI process. The study also offers implications for managers. Firstly, the framework presented allows the management of small wineries to plan activities to enact specific sommelier roles to advance along the relationship initiation process with starred restaurants from the very beginning. The wine steward could reduce the complexity of this process by playing the several roles discussed in the present study.

Therefore, this study encourages small wineries to strengthen their relationship with sommeliers. Building a relationship with sommeliers could lead to the construction of social ties and business landscape knowledge crucial for small wineries in the primary subphases of relationship initiation. Despite the benefits, small wineries are called to dedicate a significant amount of time and effort to building relationships with professional sommeliers who are in high demand from multiple wineries and subject to a multitude of stimuli. A further drawback is the complexity of communicating the wine proposition with sommeliers who, in turn, will communicate with starred restaurants.

Secondly, this study provides interesting insights for starred restaurants, suggesting how they could enhance the sommelier’s roles to fulfil their business needs and reduce the risk of relying on unfamiliar small wineries. The sommelier could help select wine suppliers and manage their relationships, allowing the starred restaurant to focus its efforts on its core business. However, starred restaurants are challenged to carefully manage their relationships with sommeliers to prevent eventually opportunistic behaviour, maybe by structuring functional contractual incentives. Also, as sommeliers are not business experts, hence, managing their role in business relationships represents a further challenge for SMEs.

Thirdly, the study emphasises the roles of sommeliers in BSRI in addition to their role in taste evaluation. Hence, the study supports sommeliers’ understanding of how they can support the interaction between small wineries and starred restaurants and how their role influences their partners’ business results. Therefore, the study offers suggestions to the sommelier’s associations to enrich the training of future wine stewards from the business perspective.

In addition, the study also offers implications for academic institutions and society. On the one hand, academic institutions should conduct more research on the role of the sommelier from the business perspective and develop advanced courses focused on training sommeliers to manage the business and marketing aspects of their profession. On the other hand, the sommelier should be more recognised and enhanced as a professional capable of connecting the production and consumption context, recognising its relevance for small wineries’ business development. Finally, as sommeliers can connect the production and consumption world, the findings suggest that these actors performing various roles might foster the adoption and acceptance of agri-food innovation policies in entrepreneurial and societal contexts.

6. Conclusion

This study investigates how sommeliers support the BSRI between small wineries and starred restaurants. The research adopted the GT abductively according to the systematic combining approach. The study’s contributions are related to the BSRI, third actor roles, and the part played by wine stewards in developing the BSRI. This study proved particularly relevant for small wineries in a rural context, as it was developed by considering the wine business landscape of the Marche region. In fact, given the importance of the sommelier in the wine business scenario, empirical contributions such as the present study are relevant for small wineries and restaurant management to address wine context complexity.

Concerning the first research question, the study argues, in line with the concept of initiator contributors, that the sommelier plays a role in initiating the business relationship between starred restaurants and small wineries. The role of the sommelier (third actor) unfolds through multiple, continuous and simultaneous roles toward both actors during the BSRI process. The study highlights how small wineries can engage sommeliers to develop new business relationships with starred restaurants.

Concerning the second research question, the various roles played by the sommelier (access provider, evaluation assistant, expectations builder, provider of concrete evidence, risk reducer, power balancer, and matchmaker) foster interactions between the actors and lead them to move toward each other. In the beginning, the wine steward leads the starred restaurant to pay attention to the small winery to fulfil its wine list deficiencies. The sommelier fosters primary interactions between business actors, making them aware of the business opportunity and assisting in building reciprocal trust. Finally, the sommelier supports the interactions between actors to create the conditions for the “final” agreement (see Table 2).

In line with the existing literature, the study confirms that the BSRI process unfolds unpredictively with blurred and recursive phases. We also describe a new role that a third actor might play during the BSRI in the case of significant power and information asymmetry between the parties involved: the power balancer. Notably, some roles of the third actors seem to be the outcome of the process, and the roles performed in earlier stages unfold as the consequence of the BSRI. This new perspective on the roles of third actors during BSRI is developed in the case of sommeliers, small wineries, and starred restaurants.

6.1 Limitations and further studies

Some limitations of this study should be acknowledged, and the major one relies upon its qualitative and explorative nature. Despite being increasingly adopted to explore network interactions, context and the evolving nature of the phenomenon of interest strongly influence the qualitative approach. On top of that, it is worth highlighting that the research context is bounded in a single region of Italy, and results might be different in other countries or business contexts.

Additional studies are called to understand whether other actors besides the sommelier play a role in the BSRI between wineries and starred restaurants. In addition, exploring the role of formalised third actors in other industries represents interesting avenues for further research. Additional studies are needed to discover how small wineries could develop strong relationships with sommeliers besides facing a lack of human and financial resources.

Adopting different theoretical lenses – such as service-dominant logic and value co-creation theories - might provide additional insights and different nuances in the understanding of the matter. Specifically, the principal agent theory could give notable insights into how starred restaurants could minimise the risk of sommeliers’ opportunistic behaviour. Finally, extending the geographical scope of the study might highlight how different dynamics emerge in other wine-producing regions.

Figures

Figure 1

Data collection and analysis process

Table 1

Industrial network approach literature on buyer-seller relationship initiation

Authors/ Phases considered to be initiation	Description of the phases
Ford (1980)	This study conceptualises initiation in “the pre-relationship” and “the early” stages. In the first stage, customers assess potential suppliers; the evaluation starts with a specific episode in an existing relationship or due to referrals. In the “early stage”, primary negotiation activities occur. Due to social, cultural, technological, geographical and time distances between the parties, it is difficult for purchasers to assess a supplier's reliability. Therefore, the buyers’ judgement is based on the sellers’ reputation as a substitute for experience
Pre-relational stage
Early stage
Dwyer et al. (1987)	In the awareness phase, the parties are involved in unilateral activities to enhance their positioning and posturing to improve their attractiveness to the other parties. The exploration phase includes search and trial activities in relational exchange. This stage is composed of five subprocesses: attraction, communication and bargaining, power and justice, norm development and expectation development
Awareness phase
Exploration phase
Batonda and Perry (2003)	The model proposed by the authors describes the initiation through the “searching process” and the “starting process” phases. In the first phase, the parties focus on search and trial for potential exchange partners that they evaluate based on economic and social aspects. The “relationship starting process” tests the goal compatibility between the partners, along with the identification of interfirm and interpersonal dynamics of networks. These phases progress in an unpredictable way. They add a dormant phase (5) in their model to account for relationships established between actors that have been in contact before
Searching process
Starting process
Edvardsson et al. (2008)	In their model, the authors conceptualise initiation with status (12) (unrecognised, recognised and considered) and forces (converter (4) and inhibitor (7)) to symbolise in contrast to the notions of phases or stages that the initiation process may stop at any time. In the “unconsidered status”, the parties have no knowledge of each other; in the “recognised status”, one of the counterparts (or both) gains awareness of the other actor, and social relations are built. In the “considered status”, the counterparts discuss the objectives of the future business relationship, exchange information and exhibit trust (15) toward each other. The transition from one status to another depends on converters and inhibitors. Converters (timing of specific activities, trust, offering and competence) can speed up or slow down the process, whereas inhibitors, such as bond, risk and image, can hinder the process to proceed or reverse
Unrecognised status
Recognised status
Considered status
Claycomb et al. (2010)	By focussing on a buyer’s perspective, the authors present four stages of relationship development. The initiation of the business relationship entails the awareness and exploration phases. In the awareness stage, buyers seek information on suppliers, and no transaction occurs. In the exploration stage, the first purchase takes place, and buyers are involved in estimating and testing goal compatibility, integrity and performance of suppliers. At this stage, buyers are still considering other potential suppliers
Awareness
Exploration
Ford et al. (2011)	In the pre-relationship stage, parties face high inertia as they evaluate the potential benefits of entering into a business relationship and quantify the efforts needed in terms of investments, adaptations and learning. In addition, parties seek to build trust with their counterparts. In the “exploratory stage,” time investment and distance (6) reduction between the parties occur. The model emphasises that relationships do not follow the order of stages in a predetermined way
Pre-relationship stage
Exploratory stage
Wagner (2011)	The author suggests adopting buyer–seller relationship development through life-cycle phase logic to study supplier development activities. The author recognises that business relationships evolve through stages (initiation, maturity and decline). In the initiation phase, the buyer and seller develop relationship-specific routines to better engage in supplier development activities
Initiation
Abosag and Lee (2013)	Based on the life-cycle models (11), the authors describe the initiation with a primary phase labelled the pre-relationship stage, in which partners hear or gain knowledge of the counterpart. Then, there is the early interaction stage, in which the parties trial and test each other to reduce future relationship uncertainties
Pre-relationship
Early integration
Valtakoski (2015)	The study explores how the intangibility of the offer and the buyer’s trust in the seller affect the initiation phase. The seller’s efforts should aim to improve his trustworthiness. Therefore, the supplier must manage compensating strategies to improve the buyer’s cognitive (2) and affective trust (1). The study also emphasises that for the seller to initiate a relationship with the buyer, the buyer's trust in the seller must exceed a certain level
Initiation
Mandják et al. (2015)	The model proposed by the authors entails a “starting situation” characterised by the simple co-existence of the actors in space and time. Potential partners are not conscious of the existence of each other. In the awareness stage, trigger issues (16) draw one partner’s attention to the existence of the other. At the individual level, the trigger issues are personal reputation, prior relations and referral; meanwhile, at an organisational level, the trigger issues are network position, attractiveness, goodwill, visibility and initiative. Trigger issues are fundamental to push the actors to start communicating because the decision to contact a potential partner is made in a situation of uncertainty. The next stage of the model, the interaction process, entails trust building from both individual and organisational levels through social and information exchange episodes. In the last stage, the relationship is established, and trust has been built at both individual and organisational levels
“Business relationship emerging flow”
Ferreira et al. (2017)	Extending the dyadic business relationship initiation analysis to the triadic realms, the authors identified the beginning of the dyadic business relationship in the matching phase in which the actors are engaged in “practice matching” which aims to align buyer and supplier processes, resources and competencies
Matching
Aaboen and Aarikka-Stenroos (2017)	These studies adopt the relationship initiation process described by Aarikka-Stenroos et al. (2018). The authors develop a model of initiation where the sequence of process elements can vary, interact and co-exist. The model entails a pre-initiation phase composed by two main process elements, the “need/opportunity”, which describes recognition activities of the actors in search of business opportunity, and the “match/attraction”, which entails the research of compatible partners. Then, the actual initiation phase contains four process elements: “accessing”, “defining exchange”, “building conditions” and “forming the future”. “Accessing” consists of all gate-opening events that push the parties to start a dialogue and mutual negotiations. In the “define exchange”, the parties define the requirements for the relationship to develop. “Building conditions” concerns the creation of the conditions to foster collaboration and trust among the parties; mutual understanding activities from both buyer–seller actors take part in this stage. Finally, in the “forming the future” stage, the knowledge of common goals and long-term potential benefits allows the parties to make strategic decisions for the relationship. The authors introduce the concept of “initiation contributors” (8), who drive the initiation process forward by triggering and advancing the focal dyadic initiation process
Aarikka-Stenroos et al. (2018)
Fraboni (2023)
- Pre-initiation
- Actual initiation
Sabatini et al. (2021)	This study provides a description of the network initiation process. In the “pre-engagement”, the firm acquires an understanding of the business network and how to engage in social contact with it. In the “initial engagement”, the firm interacts with the network's social context, but no business interaction occurs. In the “engagement” period, business interactions begin, and the firm focuses on adapting products and creating bonds and trust with the actors of the network
Pre-engagement
Initial engagement
Engagement
Klimas et al. (2023)	The authors developed a multipath development framework for interorganisational relationships. In the initial phase, the authors asserted that parties experience first contact and perform mutual compatibility tests and a formal establishment of the relationship is carried out
Initiation and initial development
Wadell and Bengston (2023)	In a turbulent business network, the authors suggest that a new business relationship could start from a resource combination previously controlled by one actor. Wadell and Bengston (2023) have found that business relationships could even start from a resource combination previously controlled by one actor. This stands in contrast to previous BSRI frameworks, which described resource combination as a step subsequent to BSRI as a consequence of increasing interdependencies (9) between the business actors' resources
Starting situation in turbulent business networks
Zafari et al. (2023)	The authors have found that in turbulent environments, business relationships may not develop following the chronological stages described by existing literature. A turbulent (17) business environment increases potential business partners' perception of vulnerability (18), the urgency to act and uncertainty about future dependencies. These perceptions lead to the formation of interemistic relationships (10) characterised by a constant search for new partners and the maintenance of a diverse portfolio of relationships. The business relationship initiation phase in a turbulent business landscape is characterised by rapid development based on swift trust (14) due to the perception of urgency. This swift trust speeds up decision processes, favouring early resource exchange between the parties. Since primary interactions, both partners show high levels of commitment (3) demonstrated through regular communications and informal adaptations in order to leverage opportunities, mitigate threats and exert control activities. The intense interactions between the actors lead to the formation of trust based on fulfilled promises, which contribute to strengthening social bonds (13) between the parties
Business relationships in turbulent environments

Glossary, concepts and related definitions based on the business relationship domain

(1) Affective trust: It is the perceived willingness of one actor to engage in actions that show special affective consideration toward the counterpart (Valtakoski et al., 2015)

(2) Cognitive trust: It is based on the rational evaluation of information about the counterpart’s product/service performance and behaviour (Valtakoski et al., 2015)

(3) Commitment: It refers to the extent level of effort and input that one actor intends to put into maintaining a close and enduring business relationship (Dwyer et al., 1987; Lasrado et al., 2023)

(4) Converter forces: Forces that speed up or slow down the business relationship initiation process (Edvardsson et al., 2008)

(5) Dormant phase: Inactive stage of the business relationship initiation process (Batonda and Perry, 2003)

(6) Distance between potential partners: Distance indicates the level of closeness between the counterparts in the business relationship. Distance encompasses various forms: social, cultural, technological, time and geographical distance (Ford, 1980)

(7) Inhibitor forces: Forces that hinder the business relationship process to proceed and reverse (Edvardsson et al., 2008)

(8) Initiation contributors: Entities that contribute to business relationship initiation (Aarikka-Stenroos et al., 2018)

(9) Interdependence: It refers to the mutual understanding by business partners that participation in their business relationship provides greater value than non-participation in it or participation in an alternative business relationship (Zafari et al., 2023)

(10) Interemistic relationship: A business relationship characterised by a short lifespan and by the emergence of high interdependence and coordination efforts between the actors from the beginning due to the need to complete tasks under time pressure. The perceived lack of time forces interestimic relationships to use proxies for trusts, such as the counterpart’s reputation and third-party assurance (Zafari et al., 2023)

(11) Life-cycle models: It refers to a conceptualisation of the development of business relationships that sees the progression of the business relationship based on life cycle stages: beginning, growth, maturation and decline (Hussain et al., 2020)

(12) Status: Distinct and rather stable positions in the business relationship initiation process which differ in terms of closeness to a business agreement. In contrast to the notion of stage, the status does not presume that the business relationship initiation process progresses following a certain order: the process may stop, reverse and proceed at any time (Edvardsson et al., 2008)

(13) Social bonds: Emotional ties that keep business partners together (Zafari et al., 2023)

(14) Swift-trust: Trust formed quickly due to the lack of time to test the counterpart’s competencies (Zafari et al., 2023)

(15) Trust: It could be defined as the willingness to take risks that one actor demonstrates in order to start a business relationship with the counterpart (Valtakoski et al., 2015). Trust could also be defined as the belief and confidence in the potential business partner’s reliability and integrity (Zafari et al., 2023)

(16) Trigger issues: Entities, aspects or factors that prompt (trigger) mutual interaction process between business actors at both individual and organisational levels during the business relationship initiation. “Trigger issues transform what is potential into what is real. Their role is very similar to the role of catalysts in chemical processes” (Mandják et al., 2015)

(17) Turbulence: It is defined as an external environmental condition characterised by continuous, abrupt, sudden and unpredictable external changes of a magnitude that threaten the core function and survival of the firms (Wadell and Bengston, 2023; Zafari et al., 2023)

(18) Vulnerability: A firm's feeling of being under threats to such an extent that it may fail to operate (Zafari et al., 2023)

Source(s): Authors’ elaboration

Table 2

Summary of the findings

Table 3

Overview of the literature contributions

Topic	Literature background	Literature contribution
Unpredictable and blurred phases of buyer-seller relationship initiation process	Batonda and Perry (2003) Edvardsson et al. (2008) Ford et al. (2011) Hastings et al. (2016) Ferreira et al. (2017) Hussain et al. (2020) Klimas et al. (2023)	The study confirms that the buyer-seller relationship process unfolds unpredictively, where phases are blurred and recursive Hence, the study highlights that buyer-seller relationship development as a progression of stages is only one of the possible trajectories of relationship evolution, and it is mainly used so far to depict and simplify a more complex process unfolding over time
Expectation development phase in buyer-seller relationship initiation	Dwyer et al. (1987)	The study provides a new perspective on the phases of buyer-seller relationships, suggesting a new order in the case of small wineries and starred restaurants. In relation to the Dwyer et al. (1987) framework, the study suggests that the “expectation development stage” could precede the “communication and bargaining stage.”
The bilateral role interpreted by the third actor in buyer-seller relationship initiation	Aarikka-Stenroos and Halinen (2007) Mandják et al. (2015) Gargiulo and Sosa (2016) Aaboen and Aarikka-Stenroos (2017) Aarikka-Stenroos et al. (2018) Pemartín et al. (2019)	The study describes the bilateral role played by the sommelier, both toward the small wineries and the starred restaurants, in the beginning of business relationships
Sommelier roles	Manske and Cordua (2005) Aarikka-Stenroos and Halinen (2007) Ruiz-Molina et al. (2010) Festa et al. (2016) Aarikka-Stenroos et al. (2018)	The sommelier performs relevant roles not only in the b2c context but also in the b2b context. In line with the literature, the study pinpoints that in the small winery and starred restaurant relationship initiation, the sommelier could play the following roles: scouter, need creator, access provider, match maker, evaluation assistant, expectation builder, risk reducer and provider of concrete evidence
The sommelier’s “power balancer” role	Aarikka-Stenroos and Halinen (2007) Reynolds et al. (2009) Alonso (2010) Kähkönen and Tenkanen (2010) Aaboen and Aarikka-Stenroos (2017) Aarikka-Stenroos et al. (2018) Siemieniako et al. (2023)	The power balancer is a new role for the third actor in buyer-seller relationship initiation. This role could be performed by the third actor when there is significant power and information asymmetry between the parties during the buyer-seller relationship initiation process
Trigger in buyer-seller relationship initiation	Edvardsson et al. (2008) Mandják et al. (2015) Aarikka-Stenroos et al. (2018)	The wine steward is depicted as the trigger that fosters the buyer-seller relationship initiation between the small winery and the starred restaurant
Third actor roles	Walter and Gemünden (2000) Aarikka-Stenroos and Halinen (2007) Aarikka-Stenroos et al. (2018) Pemartín et al. (2019)	The study provides evidence that the sommelier, as a third actor, could simultaneously perform multiple roles in the business relationships between the small winery and the starred restaurant, according to the specific subphases of the buyer-seller relationship initiation described in the framework developed in the study
Trust building due to third actor activities	Aarikka-Stenroos and Halinen (2007) Mandják et al. (2015) Valtakoski (2015)	The trust builder role ascribed to the third actor could be the result of other roles performed by the third actor. Hence, it could not be considered a third-actor role in itself

Source(s): Authors’ elaboration

Note

1. Source: https://www.quattrocalici.it/regione/marche (Access on date: 06/09/2023)