The rise of collaborative engagement platforms

Abstract

Purpose

The purpose of this study is to define and analyse the emergence of collaborative engagement platforms (CEPs) as part of a rising platformisation phenomenon. Contrary to previous literature on engagement platforms (EPs), this study distinguishes between formalised and self-organised EPs and sheds light on collaborative EPs on which heterogeneous actors operate without central control by legislated firm actors.

Design/methodology/approach

Drawing on institutional work theory, this paper explores the institutional rules, norms and practices involved in the emergence of a new platform. This paper implements a longitudinal case study of a local food network called REKO and explores how engagement practices and institutional work patterns catalysed its emergence during 2013–2020.

Findings

The findings of this study show that actors engaged within the REKO platform participated in institutional work patterns of disruption, creation and maintenance, which drove the development of the platform and ensured its viability.

Research limitations/implications

This paper encourages future research to further explore how different types of EPs emerge and function.

Practical implications

The rise of CEPs pushes the dominant managerial orientation to progress from the management “of” a platform to managing “within” a platform. For managers, this means developing novel practices for engaging and committing a versatile set of actors to nurture open-ended, multi-sided collaboration.

Originality/value

This study contributes by conceptualising different types of platforms with a particular focus on CEPs and explicating the engagement practices and institutional work patterns that catalyse their emergence.

Citation

Leipämaa-Leskinen, H., Närvänen, E. and Makkonen, H. (2022), "The rise of collaborative engagement platforms", European Journal of Marketing, Vol. 56 No. 13, pp. 26-49. https://doi.org/10.1108/EJM-11-2020-0798

Introduction

Researchers and practitioners are increasingly using the term “platform economy” to refer to novel ways of organising marketplace exchange through digitally assisted platforms (Cusumano et al., 2020; Kenney and Zysman, 2016; Marshall et al., 2016; Laurell and Sandström, 2016). In many industries, platformisation has disrupted the linear value chain logic and transformed the way businesses operate (Wirtz et al., 2019; Andersson Schwarz, 2017; Fehrer et al., 2018b; Ramaswamy and Ozcan, 2014). Despite the growing interest in platformisation, conceptual discussions of the nature of platforms as venues for exchange remain nebulous (Breidbach et al., 2014). Previous studies have repeatedly built on the concept of the engagement platform (EP), referring to “assemblages of persons, processes, interfaces and artifacts” (Ramaswamy and Ozcan, 2014, p. 34) that structure resource integration and value co-creation among their members. Thus, EPs are seen as venues that link multiple actors in the participatory process of value co-creation (Breidbach and Brodie, 2017; Fehrer et al., 2020; Li et al., 2017; Vargo and Lusch, 2017). However, more research is needed to illuminate the multi-actor nature of EPs and how actors together (e.g. both firms and consumers) design and organise platforms (Li et al., 2017).

To date, firm-centric viewpoints have dominated the extant literature; the majority of EP studies have analysed EPs that emphasise the focal actor’s (usually a platform owner company) central role in organising the platform (Blasco-Arcas et al., 2020; Fehrer et al., 2020; Hollebeek, 2019; Sarmento and Simões, 2019). Meanwhile, studies have overlooked the rise of smaller local platforms where consumers also play a central role. Some exceptions are provided by Fehrer et al. (2018a) on collaborative economies and Hollebeek et al. (2017) on virtual brand communities. In the consumer research field, Scaraboto (2015) and Mamali et al. (2018) explicated the ensuing tensions when consumer collectives grow into more formal systems, which they called hybrid economies. These types of economies operate differently from formalised platform economies because of their multi-actor nature, which further indicates greater variety in the norms, rules and values embedded in the platform operations (Bowden et al., 2017; Hollebeek et al., 2017). To further elaborate on the nascent ideas of EPs as collaborative multi-actor setups and offer an explicit distinction between firm-centric, formalised EPs and self-organised EPs, the current study introduces a specific type of EP – the collaborative engagement platform (CEP).

Building on Breidbach et al. (2014, p. 596), we define CEPs as a self-organised constellation of physical and virtual touchpoints designed to provide institutional support for the mutual exchange and integration of resources for value co-creation. Our empirical analysis draws from a longitudinal, qualitative case study of a local food network called REKO (which stands for “fair consumption”). Initiated in Finland in 2013, REKO is a network comprising local food nodes that include three actor types: consumers, food producers and voluntary administrators. The weekly exchange of local food products takes place in physical locations where consumers buy products directly from local producers. Besides face-to-face exchanges, REKO nodes are organised in closed Facebook groups in which producers announce their offerings and consumers pre-order food items (Ehrnström-Fuentes and Leipämaa-Leskinen, 2019; Leipämaa-Leskinen, 2021). Thus, the REKO platform operates on both physical (food delivery) and virtual (Facebook) touchpoints (Breidbach et al., 2014).

Further, while prior EP works have focused on conceptualising actor engagement, for instance, by revealing the effects of engagement conditions, properties and outcomes on platforms (Li et al., 2017) or how focal actors can organise engagement through orchestrating, facilitating and stimulating processes (Blasco-Arcas et al., 2020; Storbacka et al., 2016), we analyse how CEPs emerge and are maintained. By doing so, we draw on the institutional theory to reveal the processes by which structures become established and meaningful in social life (Koskela-Huotari et al., 2020; Greenwood et al., 2008, p. 33). In particular, we focus on the dynamic side of institutions and their restructuring through institutional work (Dahlmann and Grosvold, 2017; Lawrence and Suddaby, 2006). We argue that institutional work patterns:

• disrupt old institutions – the taken-for-granted roles, norms and practices of market actors in traditional market exchange (Koskela-Huotari et al., 2020); and

• build and maintain new institutions that support the development of engagement practices between different types of actors, such as consumers and producers (Dolbec and Fischer, 2015; Scaraboto, 2015) and, thus, catalyse the emergence of CEPs.

The purpose of this study is to define and analyse the emergence of CEPs by addressing two research questions:

RQ1.

How do actors carry out different types of engagement practices during the emergence of collaborative engagement platforms?

RQ2.

How do institutional work patterns catalyse the emergence of collaborative engagement platforms?

The study contributes to EP research, first, by revising and expanding the conceptual nature of different types of EPs (Breidbach et al., 2014; Frow et al., 2015; Fehrer et al., 2020). We view EPs along a continuum where central control dominates on one end and self-organisation of actions prevail on the other. Second, we add to the previous EP literature by showing how CEPs emerge from the actor interaction and institutional perspectives. Our findings reinforce prior discussions on the engagement side of EPs (Breidbach and Brodie, 2017; Hollebeek et al., 2017; Storbacka et al., 2016), as we shed light on the institutional rules, norms, practices, meanings and social relations that must be evaluated and developed to ensure the viability of the platform.

The article is structured as follows. In the next section, we present the conceptual background and theoretical framework of the study. The third section reports on the longitudinal, ethnographic research setting and data analysis. Fourth, the findings present a detailed case analysis of the development of REKO and how the actors participated in institutional work patterns during its emergence. Finally, we conclude by discussing theoretical contributions, practical implications and future research avenues.

Conceptual background

Engagement platforms in service research

Early EP research emphasised the firm-centric, purposeful design and ICT-enabled nature of platforms that enable novel ways for companies to collaborate with customers (Prahalad and Ramaswamy, 2004; Breidbach et al., 2014; Ramaswamy and Ozcan, 2014; Storbacka et al., 2016). Accordingly, EPs were defined as purpose-built, ICT-enabled environments containing artefacts, interfaces, processes and people (Ramaswamy and Gouillart, 2010). Recent discussions have adopted a more actor-centric or balanced consideration of EPs (Breidbach et al., 2014). For example, Storbacka et al. (2016, p. 3011) conceive of EPs as “multisided intermediaries that actors leverage to engage with other actors to integrate resources”. Nevertheless, scholars largely agree that systematic conceptualisations of EPs are still to evolving (Breidbach et al., 2014; Storbacka et al., 2016). Below, we elaborate on the conceptual discussions, pointing to the difference between formalised and self-organised EPs.

Most prior studies have discussed EPs provided by focal firms (Blasco-Arcas et al., 2020; Hollebeek, 2019; Sarmento and Simões, 2019). Studies have explored formalised EPs and suggested that, with proper design and activities, EPs will engage actors for collaboration (Breidbach et al., 2014). Furthermore, the physical characteristics of EPs have been identified; for example, Frow et al. (2015) identified five types of EPs that connect actors and enable their interactions:

1. digital applications;

2. tools or products;

3. physical resources;

4. joint processes; and

5. dedicated personnel groups.

Platforms have also been categorised on the basis of their primary purposes by differentiating the interactional (operating and instrumental) or transactional (enabling and supplying) platforms (Breidbach et al., 2014) or intermediaries, technology creators, matchmakers or living structure platforms (Fehrer et al., 2020). Consequently, only a couple of prior studies have analysed EPs that are established and hosted in a decentralised manner, such as online brand communities (Bowden et al., 2017; Hollebeek et al., 2017) or multi-actor platforms (Li et al., 2017). In addressing the capability of platforms to foster engagement between actors, these studies have revealed that distributed agency among the various actors entails specific challenges for platform organisation.

To address the variety of EPs, we propose a general distinction between formalised and self-organised EPs. We argue that formalised EPs operate through transactional exchange and that platform owners (usually a company) play a focal role in organising the platform. Typical examples include e-commerce (e.g. Amazon) and online service (e.g. Uber) platforms. Conversely, self-organised EPs often include heterogeneous actors (e.g. producers, consumers and administrators) that implement relational exchanges for nurturing long-term relationship development. Typical examples of self-organised EPs are online consumer communities that operate without legislated firm actors and collaborative EPs (i.e. CEPs), such as the REKO local food network, where producers and consumers share resources and administrators facilitate the operations. It is of course possible for different types of EPs to move along the continuum, as they are continuously developing – for instance, a self-organised EP may become more centrally organised as it grows larger (Mamali et al., 2018). However, for the purposes of analysis, the continuum is still a useful way to distinguish between different types and their characteristics.

In addition to the types of EPs, the engagement element of the EP concept is crucial. Engagement is regarded as a prerequisite for resource integration and value co-creation on the platform. According to Storbacka et al. (2016, p. 3009), actor engagement provides the micro-foundation for value co-creation and has been defined as “both the actor’s disposition to engage, and the activity of engaging in an interactive process of resource integration within a service ecosystem”. The notion of engagement practices (i.e. institutionalised actor-to-actor interactions on the EP; Breidbach and Brodie, 2017; Hollebeek et al., 2017) offers a means for bridging the duality between individual and collective levels of engagement. Thus, engagement practices represent the concrete interaction mechanisms and processes that manifest actor elements such as background understandings, emotional engagements and goals (Schau et al., 2009; Hollebeek et al., 2017).

In this study, we focus on engagement practices to illuminate how REKO actors interacted during the CEP’s development. We show not only how the different actor groups interacted with each other but also how the engagement practices were connected to the institutional work that drove REKO’s emergence. Although engagement practices are the concrete identifiable interactions occurring in the CEP, we use the more abstract concept of institutional work to analyse the emergence and formation of the platform as a whole.

Institutional work in collaborative engagement platforms

Institutional theory (Hoffman, 1999; Lawrence and Suddaby, 2006; Scott, 2005) has been increasingly applied to studies on the service-dominant logic (Vargo and Lusch, 2016; for a review, see Koskela-Huotari et al., 2020). These discussions draw from the basic assumptions outlining that institutions are perceived and acted upon in ways that simultaneously reproduce or alter them (Giddens, 1984). In service ecosystems, this interplay manifests in resource integration (Vargo and Lusch, 2016; Lusch and Vargo, 2014, pp. 11–12). To date, the EP studies have not used the institutional lens in exploring EPs and their operations.

To fill this gap, we analyse how actors design and organise EPs, especially through collaboration, and thus, we adopt the concept of institutional work (Thornton et al., 2012), which focuses on the institutionalisation process at the mid-range level from the viewpoint of active actors (Brodie et al., 2019; Koskela-Huotari et al., 2016; Vargo and Lusch, 2017). Institutional work is “the purposive action of individuals and organisations aimed at creating, maintaining, and disrupting institutions” (Lawrence and Suddaby, 2006, p. 215). Koskela-Huotari et al. (2016) described institutional work in service ecosystems as unfolding through actors’ breaking, making and maintaining institutionalised rules of resource integration. They also stated that, particularly in innovative service ecosystems, maintaining at least some old institutional rules is mandatory because actors must recognise and relate to the institutional arrangements. One way of engaging actors is through value resonance (Aal et al., 2016, p. 626): “When all engaged actors in a service ecosystem enact and live the same core values, resources integration will become more effective […] Shared values provide a compass giving energy and direction for joint efforts.” However, institutional work also requires time and effort as well as experimentation by service ecosystem actors. Thus, it also constitute “ongoing negotiations, experimentation, competition and learning” (Koskela-Huotari et al., 2016, p. 2966), including the likelihood of a mixture of old and new institutions (Siltaloppi et al., 2016). These institutional mechanisms are essential to understanding how CEPs emerge and how actors take part in this process.

Research framework

To build a broader theoretical understanding of the rise of CEPs, we present a research framework (Figure 1) that focuses on the interface of value chain and platform logics. The framework shows that the platform logic diffuses to areas where the value chain logic has been incumbent (Marshall et al., 2016). This co-existence of logics gives rise to the emergence of different types of platforms. For the focal study, the value chain logic is a point of departure for the launch of institutional work patterns of disruption, creation and maintenance work (Lawrence and Suddaby, 2006). This supports the development of engagement practices in steering and supporting actor-to-actor interactions (Siltaloppi et al., 2016) and, thus, catalyses the CEP emergence.

The value chain logic has long been a dominant order in organising exchange in various areas of business (Porter, 1985). Accordingly, value chain members comprise a linear order of activities through which upstream production and producers are connected with downstream consumption and consumers (Fehrer and Wieland, 2021). The value creation in such setups lies in value-adding activities whose competitiveness is determined by market forces regarding quality and cost (Porter, 1985). Digitalisation and the related increase in technology-enabled connectivity have long supported the rise of platforms exemplified by the circular (Fehrer and Wieland, 2021) and sharing economy (Ciulli and Kolk, 2019; Eckhardt et al., 2019) and platform business models (Fehrer et al., 2018b; Wirtz et al., 2019). These setups feature a platform logic according to which platforms are venues through which various type of actors may exchange resources in a participatory process of value co-creation comprising direct and indirect actor-to-actor interactions (Mody et al., 2021; Dellaert, 2019).

The circular form of the institutional work patterns in the framework communicates that the CEP represents an emergent, constantly developing setup in which continuous institutional work and parallel activities of disruption, creation and maintenance support its existence. Consequently, the present study addresses the collaborative ways of organising resources to design, conduct and maintain new modes of exchange (Mamali et al., 2018; Scaraboto, 2015) and institutionalise the new logics behind them (Kennedy et al., 2020). In particular, our empirical case provides a novel perspective to explore how the newly emerging CEP stays viable in the institutional processes where various actors and their norms, values and practices come together. The above-discussed framework provides the structure of our longitudinal empirical study.

Research method

We used a longitudinal case study of the REKO platform from its initial stages in 2013–2020. Initiated in the Ostrobothnia region of Finland, REKO provided a grassroots response to increasing local food demand. In 2020, Finland had more than 200 local REKO nodes with over 280,000 members; Sweden had more than 120 REKO nodes; and there has been increasing activity in other Nordic countries, Italy, Ireland and Canada. The selection of the case study research design was driven by our aim to advance understanding of the complex and multifaceted phenomenon of the emergence of CEPs in a real-life setting (Leonard-Barton, 1990; Piekkari et al., 2010). Furthermore, the case study provided the researchers with an opportunity to gain empirical insights and generate a better understanding of the heterogeneous actors and their mutual interactions within a CEP context (Andersen et al., 2016).

We applied an ethnographic research design (Hammersley and Atkinson, 2007), and the overall case account covered a seven-year period from the end of 2013–2020. The data collection started in 2014, and the most intensive period of the empirical study spanned between 2015 and 2016, during which we also created a retrospective account of the events, actions and interactions pertaining to REKO’s first years. The data collection ended in 2020. We focused on several REKO nodes, starting with Vaasa, the oldest in Finland (initiated at the beginning of 2014, with more than 13,000 Facebook members as of May 2020). For the other case examples, several smaller REKO nodes were selected from the Pirkanmaa region (Kangasala, Nokia, Pirkkala and Hervanta). They were initiated in 2014 or 2015, with 1,800 to 3,800 Facebook members as of October 2020.

Data collection

In line with an ethnographic research design, we used several data collection methods in different local REKO nodes to obtain a holistic and multifaceted view of the phenomenon. First, REKO Facebook group discussions were actively followed to gain familiarity with the practices and procedures of the local REKO nodes. The online observations were most intensive between 2014 and 2016, but they continued until 2020. One author followed the discussions in the REKO Vaasa Facebook group and another in the REKO Kangasala and Nokia Facebook groups. Interesting discussions and debates were documented, for example, through screenshots in field notes. The authors remained in the background and did not, for example, initiate new discussions; instead, they acted as ordinary consumer members. Because the REKO Facebook groups were closed (i.e. local administrators accepted members to the groups), we sought permission to use the Facebook quotations to report the findings.

Second, the same two authors conducted offline observations in the physical food deliveries in Vaasa, Kangasala and Nokia. To compile first-hand customer experiences, they pre-ordered food items in the Facebook groups and visited the food deliveries where they bought local food items. During the most intensive data collection period (2014–2016), the authors documented, through text and pictures, shopping experiences in field notes. In total, the offline observations lasted from 2014 to 2020. The observational data provided us with a good insider perspective on the platform’s operation from the consumer’s perspective. To supplement the observational data, the same authors also monitored REKO’s development by following related media articles until 2020.

To facilitate an in-depth understanding of the selected REKO nodes, we conducted interviews with selected informants recruited through purposeful sampling. The informant profiles and interview timings are described in more detail in Appendix 1, while the interview questions are shown in Appendix 2. We conducted 34 interviews: 20 with consumers, 10 with producers and 4 with administrators. The consumer interviews were conducted in 2014 in Vaasa and in 2016 in the Pirkanmaa region (including REKO nodes operating in the city of Tampere and the towns of Kangasala, Nokia and Lempäälä). This allowed us to monitor the consumers’ roles on the platform longitudinally along the chosen timeline. Moreover, the timing of the producer interviews differed: the Vaasa producers were interviewed in 2015 and the Pirkanmaa producers in 2016. The four administrator interviews provided the most detailed narratives and were conducted throughout the interview period. Local administrators and the founder of REKO furnished us with versatile information regarding the whole platform’s emergence and development. The interviews lasted between 30 and 90 minutes, and they were all recorded and transcribed, resulting in 500 pages of text. The informants were identified with numeric symbols in the findings section to guarantee their anonymity.

Data analysis

The data analysis proceeded according to Spiggle’s (1994) five phases of qualitative analysis: categorisation, abstraction, comparison, dimensionalisation and integration. Before the actual analysis, we familiarised ourselves with the various data sets by carefully reading the materials. We ensured that the interpretations were triangulated as two of the three authors carried out the analysis, and all three authors discussed the interpretations to achieve a consensus. The analysis started by categorising (i.e. classifying and coding) the units of data into empirically grounded meaning structures (p. 493). We coded the underlying value opportunities, rules, principles and values embedded in the data, noting that REKO actors held diverse and even contradictory views about the newly emerging platform. For example, producers pondered product pricing; consumer actors puzzled over how to trade in physical food delivery; and administrators debated the moral and operational principles of the platform (e.g. how many producers should be included in one REKO node). These initial notions led us to focus on understanding how heterogeneous actors operate on the platform and the role of collaboration in this process.

Next, we moved to abstracting, where we collapsed categories into higher-order conceptual constructs (p. 493). We used the above-outlined theoretical framework and its conceptual elements (institutional work and engagement practices) and combined the initial collaboration categories into fewer groups. During the comparison phase, we explored “the differences and similarities across incidents within the data” (p. 493). We compared the data generated from the different REKO nodes and different actor groups while also collecting additional data to supplement the interpretations. Regarding dimensionalisation, we further explored the properties of the categories, that is, the conceptual dimensions that varied empirically in the data (p. 494). The coding categories were then developed into final categories of engagement practices and institutional work patterns.

Finally, in the integration phase, the identified themes were integrated with theory to produce novel insights of the phenomenon. Following the premises of axial coding, we specified the conditions of the emerging categories (pp. 494–495). We then sketched a detailed description of REKO’s development phases and connected them with the identified work patterns to build a comprehensive understanding of the role of institutional work in catalysing a CEP. As the current analysis proceeded in an iterative manner, we moved back and forth between the phases of analysis. For example, we went back to the original coding regarding collaboration when we decided to better highlight the engagement practices in the analysis.

Findings

Initiating the REKO platform, 2013–2014

Before 2013, Finnish farmers’ direct selling of food to consumers was limited to a few active marketplaces and to a few specialty shops located on farms. This attracted high transaction costs for consumers who wanted to purchase local food. Despite the scarce supply of local food, there was burgeoning demand for locally produced food among Finnish consumers, thus providing fruitful ground for a new model to emerge.

REKO was invented by Thomas Snellman, a farmer who had long been working to improve the status of organic food in the Finnish retail market. REKO started to evolve after Snellman visited France’s local food markets in 2012 and saw how local farmers sold their produce to consumers directly from their car trunks. After his trip, he contacted a few food enthusiasts in Finland, and a small group began the process of further developing REKO. This involved the recruitment of both consumers and producers to help sketch out the system, and it soon became apparent that consumers were interested in trying out this new system, as they took on key roles in the platform’s development. An open meeting was later organised, where volunteer consumers were recruited to sign up as administrators for the local nodes. Facebook was chosen as the digital interface for the local nodes to decrease the producers’ financial risks. When the consumers placed their pre-orders through Facebook, the producers could estimate their sales more accurately than in traditional marketplace trade. Thus, the virtual touchpoint brought increased dialogue, access and transparency – central characteristics of EPs (Ramaswamy and Ozcan, 2014) – to REKO. Using Facebook also allowed the network to function without membership fees and evoked a sense of trust among the actors:

It is this sort of a good meeting place, and at the same time, as the producers tell customers about themselves and show their faces, it makes you trust them. When you are on Facebook, then you can also see what your friend has bought, what your neighbor has bought, and you are member of a social community. A very large part of this success story is related to Facebook because I think it would not have succeeded at all without it. (T15, founder of REKO)

Therefore, the decision to use Facebook as the virtual touchpoint for the system (Breidbach et al., 2014) allowed more people to engage with the platform and facilitated resource sharing and coordination (including time, effort and social resources). Soon after the initial meeting, two pilot nodes were established in Pietarsaari and Vaasa, with a small number of local actors engaged with the new platform. At the beginning, deliveries of pre-ordered food took place every fortnight in a free parking area at a defined time. This meeting operated as a physical touchpoint for the platform, allowing the exchange of concrete resources, money and food purchases and facilitating social engagement through face-to-face meetings, interactions and a friendly atmosphere. Together, the virtual and physical touchpoints provided increased resource density for the platform (Breidbach and Brodie, 2017), which is typical of sharing economy platforms. At the beginning, the participating producers wanted official agreements with the consumers to guarantee a minimum number of paying customers. This practice enabled initial engagement from both consumers and producers, but after only two months, the producers decided to forego the written agreements because of an increase in consumer membership. As they became more engaged with the new platform, the producers also noticed that they could achieve higher margins for their products and establish direct contacts with their end customers:

We were able to get producers involved through it [the agreement], but then when two months had passed, the producers let us know that there was no need for the agreements anymore because they were discussing products all the time through Facebook when questions arose. So they just announced that they no longer want the agreements and that they can just report what they are offering for each meeting. (T15, Founder of REKO)

This form of organising without contracts significantly lowered the transaction costs (especially contracting costs) in REKO exchanges (Fehrer et al., 2020). Indeed, avoiding bureaucracy was a guiding principle in the initiating phase, and actors decided on a bottom-up organisation for the platform, with only a few regulative principles and without a legislative focal actor, as in the case of hybrid economies (Scaraboto, 2015; Mamali et al., 2018). The initial REKO rules were:

• producers could only sell their own products directly to consumers;

• no extra charges were allowed within the system; and

• producers were both legally and commercially responsible for their sales.

The initial rules emphasised the disruption work of REKO actors within the prevailing structure of grocery retail in Finland. The original motivation was to decrease the number of food wholesalers, distributors and retailers in the local food distribution chain. Supporting and interacting with local farmers and voting against the major retail chains were key consumer motivations for joining the network:

I’m quite annoyed about the food monopoly position of the [two biggest grocery retail chains]. […] Producers should get a bigger share of the margin, and I am willing to pay. For instance, at REKO, price is not a factor for me, but I pay for the ability to meet the producers and to have a chance to chat with them. (T2, Consumer, REKO Vaasa)

In sum, in the initiation stage, the consumers’ engagement practices included enrolment and exchange. These involved signing up to the Facebook groups and making regular orders, thus engaging in the concrete exchange of money and goods with the producers. The following quote exemplifies this:

So I have followed my own contacts on Facebook. I’ve witnessed them taking their friends to the events and then, after which they will become members […]. And then when you share a photo of the kind of food that you’ve bought from REKO this time, there’s usually a discussion related to that. So it spreads among circles of friends, too. (T34, Administrator, REKO Pirkanmaa)

The producers’ engagement practices included enrolment, with local food producers signing up for the REKO groups and advertising their products on Facebook. They also included governing practices based, first, on formal written contracts, which were later deemed unnecessary with the emergence of trust and social governing. Finally, the administrators’ engagement included the enrolment practices of signing up as administrators, setting up new REKO nodes and regulating practices (Hollebeek et al., 2017) related to providing rules and guidelines for the platform. Word of the new platform spread through the media and word-of-mouth, and new local REKO nodes were established throughout Finland. Consequently, REKO grew exponentially during the first two years, with 27 local nodes and 12,000 members by the end of 2014.

Building the REKO platform, 2015–2016

As the number of REKO nodes and members increased, they became more heterogeneous, and the initial regulative principles were contested. This typically occurred when new local nodes were established, and administrators had to decide which producers and how many would be allowed to sell their products in the nodes. To clarify the situation, a new institution was established in 2015, a Facebook group called “REKO family”. This is an invitation-only group for local administrators to help them share experiences and negotiate REKO’s rules and principles:

A little over a year ago, when the new REKO nodes started, I discussed with [another active administrator] that we should perhaps now set up this REKO family. It is also a Facebook group, but meant for administrators […]. There, we have discussions with the administrators. If somebody has a problem or challenge, then they can ask how others have dealt with it […]. What kind of producers are accepted, what are the principles, and for instance there’s been these cases that are borderline against the rules, like if there’s a third party in-between or something. (T33 Admin, Pirkanmaa)

Establishing “REKO-family” increased communication and strengthened the relationships between the administrators. Further, at the end of 2015, a national face-to-face meeting with the administrators was organised to discuss whether REKO should be formalised as an official organisation with a legislative focal actor. However, the group decided to continue as a self-organised platform where local nodes could decide their own rules as long as they did not violate the three original regulative principles.

The debate continued in a few REKO Facebook groups following the national meeting. Also, national media reported the debate. A few administrators demanded that REKO nodes should only sell organic food because the official criteria for organic production would help redefine REKO within the context of ethical food. However, most REKO actors thought that there were no objective criteria for defining ethical or local food and that limiting the supply to only organic products would limit consumers’ ability to make their own decisions regarding which producers they wanted to support. The founder asserted that, despite his personal views, he supported local decision-making by administrators, even though conflicting situations may ensue. Below, he describes how the vision of boosting organic food production guided the foundational process at the beginning and how he realised that an overly restrictive (i.e. only organic produce) model would not work in the long run:

Because “I myself have been an organic farmer” for 25 years, I thought that this was the starting point. But I knew right away that if we restricted it to organic products, it would be hard to have an adequately wide selection. So we decided right at the start that we would try to find “smaller producers with a good reputation”. (T15, Founder of REKO)

Consequently, supporting local farming and the vitality of Finnish agriculture was deemed more important than limiting REKO to organic production: focusing on organic food would be overly restrictive within Finland’s geographical and social context, where agricultural conditions vary considerably according to geographical landscapes and consumer demand. Furthermore, the REKO consumer interviewees reported that, while they tried to buy organic food as much as possible, it was not always possible because of limited supply or high prices. In such cases, buying local was often perceived as the second-best alternative for enacting their food-related values. To illustrate, organic food production is more common in the southern and western parts of Finland than in the eastern and northern parts. Later, the consumer-administrators needed to balance the recruitment of enough producers for the nodes with ensuring that the farmers’ production methods met REKO’s ethical standards:

We try to get “as much organic produce as possible”. We have, depending on the node, I would say two-thirds of our sales products are organic. But then, for instance, when it comes to animals, being organic is not the requirement – but the fact that they are produced locally. For it to be both local and organic, that often seems to be an impossible combination. There is just “not enough supply”. (T33, Admin, Pirkanmaa)

In the spring of 2016, a few administrators decided to break from the network and create their own FM-REKO nodes (where FM indicates farmer’s markets), which only sold products with the official organic produce trademark. However, most REKO nodes remained unchanged, with decisions regarding which producers to accept made on a case-by-case basis:

For instance, one importer of coffee informed me that it is cooperating with this smaller producer in France from where it then buys the fair-trade coffee. Now, it would like to use REKO because it thinks that this kind of operation fits its principles. […] So far, “we have not accepted these types of producers”, even though they would, in principle, meet many of the criteria […] I don’t know the detailed conditions of acceptance [for individual nodes], but I know that I would not accept some of the decisions being made. But “I do not want to be the police” in that way. (T15, Founder of REKO)

Thus, in the building stage, the administrators’ engagement practices included co-operation and governance as they set up a new “REKO family” group to strengthen their mutual relationships and help each other with problems arising at the local level. Consumers used social networking practices (Schau et al., 2009; Hollebeek et al., 2017), and enthusiastic REKO members began spreading the word about the new network to their friends and family and recruited new members for the platform. In contrast to brand communities (Schau et al., 2009; Hollebeek et al., 2017), the social networking practices during this stage of REKO’s development were primarily aimed at recruiting non-members to the platform. Furthermore, producers carried out co-operative practices by assisting each other with food deliveries and marketing. These resembled the practice identified in brand communities as “assisting” (Hollebeek et al., 2017) but were more akin to concrete help in REKO. For instance, one producer could deliver customer orders on behalf of another producer if they were unable to attend the REKO meeting. The producers also engaged in mingling practices (Hollebeek et al., 2017). They made friends with each other, highlighting the platform’s collaborative rather than competitive nature:

Then if there’s sometimes something left [after the REKO delivery meeting], we give each other these products as a token of friendship. And we use each other for taste testing. But of course we also interact a lot with each other, so I have made many new acquaintances here. There’s a couple of others like me that I usually chat with – we are like-minded. (T13, Producer, Vaasa)

Sustaining the REKO platform, 2017–2020

The REKO platform continued its growth to a nationwide CEP with more than 230,000 members and 160 local nodes across Finland by the end of 2017. However, even though there was a large Facebook membership, the number of paying customers remained low, even decreasing following the initial enthusiasm. Thus, most of the consumers appeared to be more engaged with the traditional retailing system than with the new platform. One administrator initiated an active discussion on the decreasing number of consumers in the REKO Vaasa Facebook group at the end of 2016, which remained more or less active for over a year. All the actor groups actively participated in the discussion, but most were consumers. There were two contrasting viewpoints in the discussion. The first supported producers and “blamed” consumers for being used to getting superior service and variety in supermarkets:

You can notice it from this chain of postings that the “consumers are used to having everything perfect”: everything on the same shopping trip, for a cheap price, and right when you yourself want it. But supermarkets serve that purpose […]. Perhaps those people are not the most important target group for REKO. “I myself value quality and responsibility” in production over other things. In REKO, we could emphasise the fact even more that the products are unique (in REKO, I also avoid products sold in supermarkets), and they are always fresher than in a store. REKO represents responsible retailing without middlemen, wholesalers, or retailer interests. “I am ready to pay for these things”, for organic food and freshness, at the set price, and I also think that the events are well scheduled (I can drive there directly from work). (Consumer, REKO Vaasa Facebook, February 2017)

Thus, even though the costs may have been higher compared to traditional supermarkets, committed REKO consumers referred to value resonance (Aal et al., 2016) and were willing to accept higher costs to support the system. Thus, strongly verbalising their commitment to REKO could be considered important at this stage. The other standpoint addressed the importance of customer service and marketing, attempting to identify how to better engage consumers on the platform. Consumers were considered powerful actors, and marketing was considered a necessary practice that involved learning how to treat customers (Chipp et al., 2019). For instance, discussion topics included ways to shorten queueing times and how to organise delivery times and places for consumer convenience. Making purchasing easier would reduce consumers’ transaction costs, leading to positive network effects (Fehrer et al., 2018a). Also, much practical advice was offered to producers:

For me, the time of the collection event is sometimes a problem. Many work until 5:00 pm or later. But “many things suggested here could help”. Notify earlier if you are going to participate in the event, have a clear REKO sign, and sometimes, arrange a pop-up REKO at a regular market to “reach the news threshold and find new customers”. [The local parking facility provider or tourist office] can surely arrange free parking for these events if it is only for that one time. REKO has had a lot of publicity after all. Then also “producers could think of a bonus system”. For instance, every 10th visit, you could receive a coupon worth, for instance, 10 euros that you could spend freely on something from any of the stalls. “Think big and quickly”, everything may not be feasible, but we must start somewhere. (Consumer, REKO Vaasa Facebook, December 2016)

This consumer introduced familiar institutions from a grocery retailing context into the discussion, including rewarding customers for their loyalty as a possible engagement practice for producers. In the end, a vote was taken on delivery time and place to make the experience as hassle-free as possible, which resulted in splitting deliveries into two locations, one after the another, to make it easier for the producers. Because of this discussion, some producers also began using queue numbers to speed up delivery.

Even in the sustenance phase, regulative and normative principles remained core discussion topics. For instance, the optimal number of producers within one REKO node was constantly under negotiation, showcasing the imbalance between consumer and producer roles (Scaraboto, 2015). Producers needed a sufficient number of customers, but consumers wanted greater variety and price competition. A few producers left the node because they saw the high number of competing producers as unfair. In general, Facebook offered a virtual touchpoint that allowed open discussion regarding fair pricing and ethical production. These discussions became more popular after the platform was established and members became more engaged with each other. Below are examples of posted comments of a Facebook discussion chain (April 2019) in which several consumers and producers discussed pricing strategies:

Consumer: I think that the prices are so high! I understand that it costs more to farm ecologically, but there are no middlemen who should be paid, no store rents or staff costs, so “why such high prices”?

Producer: A small “producer pays everything by themselves”, including butchering, cutting the meat, packing it, the packaging materials, freesing it, transporting it back to the farm, and storing it before it is sold to the customer. Also, the costs of the waste are met by the producer. Then there’s the transportation to REKO and selling it there, paying for the card transactions to the bank, and 14% of value-added tax. Then all the marketing materials are paid for by ourselves. The cold containers we pay for by ourselves. The packaging given to the customer is paid for by ourselves. The costs of monitoring the food production is paid for by ourselves, since the meat is being stored at the farm. Here’s how the price is actually formed.

These comments show that debates about REKO’s institutional arrangements involved comparisons with traditional grocery retailers and regular employment and entrepreneurship, with continuing pressure to compare the differences between REKO and the traditional system. In sum, the consumer engagement practices comprised advising, with consumers providing help and ideas for producers and administrators in an effort to develop the platform. This was considered necessary as keeping members engaged was becoming challenging. Furthermore, some consumers verbally voiced their commitment through appreciating practices (Hollebeek et al., 2017), while others disengaged, lost interest and discontinued their REKO purchases. The producers’ and administrators’ new engagement practices included negotiation, whereby they openly discussed their pricing with other producers and consumers. Also, the number and types of producers included in REKO were debated at this stage in an effort to maintain balance between consumers’ desire for variety and producers’ desire to have an adequate number of customers:

Two [producers] are quite okay, but if there are “too many [producers]”, then of course the number of orders will be reduced quite a lot […]. It is not in “my interest” anymore if only a few orders are made. Then all the income goes into gas expenses. That’s it. (T12, Producer, REKO Vaasa)

Finally, the administrators also engaged in practices of celebration (Hollebeek et al., 2017), alluding to REKO’s success in media interviews as well as through celebrating the milestones of individual nodes (e.g. anniversaries and member counts) and helping to establish a sense of importance and continuity.

Institutional work patterns catalysing REKO’s emergence

We continue the analysis at the meso-level by discussing how the institutional work patterns of disruption, creation and maintenance work catalysed REKO’s key development phases (Lawrence and Suddaby, 2006; Koskela-Huotari et al., 2016). Table 1 illuminates the most typical institutional work patterns and their manifestations in the current case study.

Disruption work aims to disconnect the newly emerging CEP from prevailing beliefs and assumptions regarding the conventional retail system (Lawrence and Suddaby, 2006). In the current case, both farmers and consumers were familiar with the conventional retail system; therefore, REKO’s founders needed to dismantle actors’ assumptions. This was accomplished through boundary work (see A1–A3 in Table 1), where actors disrupted institutions by manipulating their social and symbolic boundaries (Lawrence and Suddaby, 2006; Zietsma and Lawrence, 2010). In the initiation phase, boundary work problematised the conventional retail system’s social (economic, physical and political) boundaries. In practice, REKO was described as a new platform that advanced producers’ economic well-being and enabled consumers to support local farmers; the conventional retail system was criticised for being faceless and as underrating farmers’ significance in the food distribution chain. The criticism was also directed at the structure of the Finnish retail market as the most centralised in Europe, where only two retailers have controlled over 80% of the market for decades (Nielsen, 2019). These statements were declared both in the initial meetings and in media commentary. Furthermore, the exchanges transitioned from physical stores to Facebook and subsequent face-to-face meetings.

Later, boundary work targeted the social and symbolic boundaries within the REKO platform, as actors committed to redefining and reconfiguring REKO’s moral, socioeconomic and cultural boundaries (Lawrence and Suddaby, 2006). This included actors’ views on how to share the moral responsibility of the platform’s viability, how to share the economic resources and how actor roles should be divided (Dahlmann and Grosvold, 2017). For instance, consumers had to leave their roles as passive customers and assume more active platform member roles. Farmers had to leave their competition orientation and own-business focus to assume more cooperative roles with each other. In particular, profit-making and commercialisation remained at the core, and producers pondered whether REKO should institute more professional marketing and ordering systems. However, these suggestions were abandoned because they would have required more work from administrators and/or member payments. In the sustenance phase, when the number of REKO members continued to increase, boundary work was needed to tackle the question of how many producers should be involved in one local node. Some REKO producers thought that restricting the number of producers in a node would present better business opportunities for them. However, the majority of REKO actors believed that its success should be shared among all actor groups. Consequently, REKO’s founder highlighted the role of loss-leader products, indicating that more producers should focus on these products to increase the viability of the whole platform. Thus, boundary work also included defining the concrete material boundaries for the platform.

Another type of disruption work comprised decreasing perceived risks (see A4–A6 in Table 1). At the beginning, one of the founding group’s biggest challenges was to decrease farmers’ perceived risks in relation to non-visible customers. Finnish farmers were used to the conventional retail system where the retailer was the visible customer, and they did not believe that a sufficiently large group of individual consumers would be willing to buy food directly from producers. To overcome the producers’ doubts, Facebook was adopted to present a virtual platform touchpoint, allowing open communication between consumers and producers. This reduced producers’ financial risks because they could estimate the actual demand for their products on the basis of pre-orders. Similarly, through pre-orders, customers could ensure they would receive the products they intended to buy. In REKO’s later development, this type of institutional work assured benefits to the members of the new platform. In the building phase, when there was conflict between actors supporting more top-down regulation (FM-REKO) and those supporting bottom-up governance, the benefits were made visible in concrete examples of low numbers of organic farmers and the conclusion that this would eventually decrease the number of products available in local nodes. In the sustenance phase, institutional work was needed to illuminate the benefits of the institution with regard to actor roles, for example, how to inform consumers of the benefits to draw them to the platform.

Creation work patterns illuminate the creation of a normative foundation. One type of creation work included defining and setting the rules, principles and norms (see B1–B3 in Table 1) (Lawrence and Suddaby, 2006). In REKO’s initiation phase, the founding group defined the initial rules, principles and norms for the newly emerging platform, which were closely linked to the CEP goals. At its inception, REKO had two goals: developing a new platform to enhance local farmers’ livelihood and facilitating broad change in agricultural and food distribution practices in Finland. The duality of the initial goals paved the way for further rule negotiation and reconstruction. The initial rules left too much room for interpretation, and local administrators were increasingly faced with questions and conflicts that they needed to resolve. It was also unclear who was responsible for resolving the unclear questions. Thus, in later phases, the rules were clarified and reconstructed by explicitly deciding that locally produced food was preferred over organic food. Overall, using Facebook as REKO’s virtual touchpoint allowed the actors to openly negotiate the norms and principles.

Alongside the defining work, the actors needed to engage in identity work (see B4–B6 in Table 1), which is closely intertwined with educating oneself and learning new practices to take on a new role (Dolbec and Fischer, 2015; Koskela-Huotari et al., 2016). In the initiation phase, the identity work included a search for identity, which involved administrators taking on new roles as REKO governors and farmers embracing their new role as direct-to-consumer sellers instead of working with merchandisers. This included learning how to price, market and sell their products. Similarly, consumers needed to learn how to trade properly by placing pre-orders through Facebook and learning how to pick up the products. In the building phase, when the internal conflict laid bare REKO’s ambiguous moral boundaries, the administrators established a new institution, a “REKO family” Facebook group to legitimise their platform roles (Zietsma and Lawrence, 2010). In the sustenance phase, most of the identity work targeted realigning actors’ identities through a balance between consumer and producer responsibilities. Identity work pointed to the problem of free-riding consumers in the Facebook groups. Our findings show that producers became frustrated with consumer members who did not actually make REKO purchases, even though they were members. Indeed, many consumers seemed to regard REKO merely as an alternative outlet for food purchases, reaping the REKO-related benefits without any responsibility for maintaining the CEP. Previous studies have also indicated that it can be difficult for consumer members to engage with alternative food systems (Leipämaa-Leskinen, 2021). Similar findings have been found in other contexts; for example, Kaptein et al. (2015) showed how increased engagement in an online platform actually led to decreased spending.

The third type of creation work involved mimicking (see B7–B9 in Table 1), where actors associated the existing practices, technologies and rules with the new institution (Lawrence and Suddaby, 2006). With REKO, mimicking first targeted the existing local food networks, such as when producers contracted with consumers following principles of community supported agriculture (Thompson and Coskuner-Balli, 2007). In addition to other local food networks, actors mimicked conventional retail practices by imitating retail prices when pricing their own products. In the later phases, mimicking work allowed REKO to expand as the new nodes and their administrators followed the rules and practices tested in the more established nodes. In future, we believe that mimicking successful local food initiatives may help REKO actors identify better ways to organise the platform.

Third, the case study revealed the importance of maintenance work, whereby actors support, repair or recreate the social mechanisms that ensured platform compliance (Lawrence and Suddaby, 2006). Thus, maintenance work involves identifying the CEP’s eligible rules and policies (see C1–C3 in Table 1). In the initiation phase, REKO actors developed eligible rules to ensure the emergence of the new platform. They purposely developed as few primary rules as possible to avoid bureaucracy and top-down regulation, believing that this would support the new platform’s social mechanisms. However, our findings showed that the transition from the platform’s initiation phase to its sustenance phase was challenging because REKO’s primary rules encouraged members to freely decide their level of activity on the platform. It was not possible to make official agreements between consumers and producers, which evoked a regular debate regarding free-riders and how to better engage consumer members within the platform. Therefore, in the phases that followed, the actors explicated and strengthened rules and policies to better integrate REKO’s core values with general social values. This work resembled the values resonance work discussed by Aal et al. (2016).

In addition to policy work, maintenance work included valorising work patterns (see C4–C6 in Table 1). At the beginning, REKO’s founder dedicated his time to spreading the word about the new platform and describing REKO’s potential. Later valorising work clarified REKO’s boundaries, particularly in relation to distinguishing the FM-REKO nodes from the original CEP. In the sustenance phase, valorisation highlighted REKO’s rise against the broader agro-food system and how REKO was an agent in moving the conventional system toward more sustainable and socially just practices. As concrete examples of valorisation, the founder received both national and international prises for his efforts in developing the CEP, which raised REKO’s profile in the media.

Maintenance work also included embedding and routinising REKO into the actors’ normative foundations (see C7–C9 in Table 1) by transforming the familiar normative foundations into the newly emerging system. For example, our findings showed that REKO was seen as embodying the traditional and nostalgic meanings of the marketplace. Later, embedding work was actualised through emotional bonding and celebrations, where actors strengthened emotional bonds by thanking each other, posting positive experiences, sharing recipes and organising campaigns to increase the demand of seasonal food. Milestones, such as REKO node birthdays, were celebrated in both the virtual and physical CEP touchpoints.

Discussion

Theoretical contributions

This study answers the call for further empirical research on EPs and their characteristics in an era of rising platformisation (Breidbach et al., 2014; Fehrer et al., 2020; Storbacka et al., 2016; Wirtz et al., 2019). It provides three main theoretical contributions. First, through the in-depth case study of a local food network, we revised and expanded the EP concept by providing a discussion of a particular type EP, the CEP, to characterise self-organised platforms that lack a dominant firm and centralised control. The extant research has focused on formalised EPs, addressing EPs that focus on new technological solutions, intermediation, brokerage and matchmaking between actors (Fehrer et al., 2020) and EPs that enable firm–consumer interactions and revenue generation (Brodie et al., 2019). While these conceptualisations do not cover the rapidly increasing forms of local collaborative platform economies, we suggest focusing on CEPs to better grasp how both firms and consumers together design and self-organise the physical and virtual touchpoints to create and maintain a platform.

Previous EP research has found that EPs enable multiple stakeholders to connect and interact; thus, they are not closed systems but operate transparently through access, dialogue and reflexivity (Ramaswamy and Gouillart, 2010; Ramaswamy and Ozcan, 2014). Nevertheless, the literature paints a rather straightforward, effortless and positive picture of EPs, suggesting that, with proper design, an EP will facilitate a company’s ability to collaborate with its customers. Thus, our second contribution highlights the important role of engagement practices to catalyse the CEP. Our results alluded to continuous negotiations, debates and adjustments, which have implications for actor-to-actor interaction within the CEP. While a previous study (Hollebeek et al., 2017) identified these practices in the context of a brand community, we presented further insights into how various engagement practices, such as social networking, celebrating and governing, manifest in the context of CEPs. Furthermore, even though CEPs have increased resource density (Breidbach and Brodie, 2017), different actors’ resources may not be easily aligned and integrated because of disagreements over actors’ rights and responsibilities in the CEP. The effort to resolve these disagreements and ensure that the CEP stays viable requires more effort than what is commonly expected of regular platform users in formalised EPs.

Finally, our study presents novel insights from institutional theory and the EP literature. We identified institutional work patterns of disruption, creation and maintenance, which catalyse CEP emergence. The extant EP research has largely concentrated on actor engagement on platforms (Li et al., 2017; Storbacka et al., 2016), without paying attention to platform emergence and maintenance. Our nuanced empirical analysis identified institutional work patterns in the three phases of CEP development (initiation, building and sustenance). Consequently, the present study observes that although a platform’s virtual and physical touchpoints may provide the basic infrastructure for collaboration, they are not enough to ensure actor engagement with the platform (Kaptein et al., 2015). To avoid an overly positive view of the relationship between actor engagement and platform viability, we argue that the institutional rules, norms, practices, meanings and social relations must be constantly evaluated to ensure positive network effects through platformisation (Fehrer et al., 2018a).

Practical implications

Our study provides practical implications, particularly for B-to-C companies, by facilitating managers’ consideration of how to position the company in regard to CEPs. From the focal company viewpoint, a key managerial task lies in sensing opportunities regarding CEPs, evaluating them with regard to company resources and their intended presence in consumer markets. Thus, company stance is not an issue of digital presence but a strategic orientation that requires either positioning the company in spearheading development in terms of adopting a proactive orientation among its industry members or deciding to be an active follower. For instance, in the context of the current study, some traditional grocery retailers in Finland volunteered their large parking lots as free spaces for REKO platform meetings, hoping to entice REKO customers to also visit their stores. Others attempted to include more locally sourced food in their selections to remove the need for customers to use REKO. However, the strategic decision in choosing between proactive and reactive platform strategies is crucial and needs to be accompanied with a plan of how the opportunities in platformisation are followed, evaluated and implemented, including a clear schedule and the roles and responsibilities of dedicated key persons.

As CEP actors (e.g. consumers and producers) occupy largely equal roles, management is more about managing “within” a platform rather than management “of” a platform. Thus, the company should set strategic objectives regarding the role of the platform in its business and aim to constantly sense and seize the opportunities brought about by the collaboration. First, a company should set clear objectives regarding operations over which it has more control. For example, a company may see the platform as a venue for learning and a laboratory for testing and innovating to gain consumer insight and knowledge on weak signals, trends, practices or some specific product features. Second, a company should leverage the platform in its communications and branding by linking the platform values into its brand identity to create positive brand associations. Nevertheless, as the purposes and activities are only partially controlled by a company, there are risks involved. A company should carefully evaluate and monitor fits and misfits regarding the values of both the platform and company and whether the benefits of the CEP may be transferable to support other branches of the company business. This fit and leveraging between the company operations on the platform and elsewhere are a strategic issue and should be resourced accordingly.

In terms of societal implications, we conclude that CEPs provide more opportunities for consumers to participate in collective actions that they deem important. In the current context, these relate especially to social and ecological pillars of sustainability, example supporting small food producers’ financial status, ensuring local food production and shortening the food distribution chains (Leipämaa-Leskinen, 2021). Thus, CEPs may provide ethical and sustainable alternatives for dominant market systems.

Limitations and future research

One limitation of the study is that, by focusing on a single case study, we were unable to compare our findings with similar cases. To determine whether the institutional processes and their dynamics are similar across cases, more research is required. The identified engagement practices and forms of institutional work could also be studied in different contexts and with different methodological approaches to determine their similarities or differences.

In general, we encourage future research to explore different types of CEPs further. Comparing our findings with empirical cases where institutional work has failed and the CEP terminated, it would help identify the factors associated with successful CEPs. Furthermore, we encourage researchers to broaden their consideration to all types of EPs through the lens of institutional theory. For instance, research could adopt the concept of institutional logic at a more macro level and focus on how different institutional logics overlap and combine in the emerging platform and sharing economies. Focusing on the impact of CEPs on broader market-level dynamics would also be a fruitful research exercise. For instance, Finland’s REKO case has increased interest in local food marketing by traditional grocery retailers, which can be interpreted as the CEP’s influence on the broader retailing system. Future research could also investigate the influence of CEP engagement on actors’ (e.g. consumers and producers) beliefs, expectations and practices in the longer term.

Figures

Figure 1.

Theoretical framework

Table 1.

Institutional work patterns alongside the development of REKO

Institutional work patterns	Initiating the REKO platform	Building the REKO platform	Sustaining the REKO platform
Disruption work targeting beliefs and assumptions	A1. Problematising the social boundaries of the conventional retail system	A2. Redefining the social and symbolic boundaries within REKO	A3. Reconfiguring the social, symbolic and material boundaries within REKO
	A4. Decreasing REKO's perceived risks	A5. Assuring members of REKO's benefits	A6. Reassuring members of REKO's benefits
Creation work targeting normative foundations	B1. Defining the rules, principles and norms	B2. Negotiating the rules, principles and norms	B3. Reconstructing the rules, principles and norms
	B4. Searching the identities	B5. Legitimising the identities	B6. Realigning the identities
	B7. Mimicking the retail actors and the existing local food exchange models	B8. Mimicking the more established REKO nodes	B9. Mimicking the successful local food exchange models
Maintaining work targeting social mechanisms	C1. Finding the eligible rules and policies	C2. Explicating the rules and policies	C3. Strengthening the rules and policies
	C4. Valorising REKO's potential	C5. Valorising REKO's social and symbolic boundaries	C6. Valorising REKO's rise within the conventional agro-food system
	C7. Embedding REKO into the actors’ normative foundations	C8. Emotional bonding between the REKO actors	C9. Celebrating the achieved milestones

Table A1.

Informant profiles

REKO Vaasa
Consumers (gender)	Age	Occupation	Family status	Thread no.	Year
A (female)	37	Student	Husband and two children	1	2014
B (male)	44	Journalist	Wife and three children	2	2014
C (female)	34	Researcher	Husband and one child	3	2014
D (female)	36	Engineer	Husband, no children	4	2014
E (female)	37	Midwife	Husband and two children	5	2014
F (male)	44	Journalist	Wife and five children	6	2014
G (female)	37	Sales manager	Husband and two children	7	2014
H (female)	24	Student	Single	8	2014
I (female)	54	Cultural worker	Single, grown-up children	9	2014
Producers	Products
Egg producer	Free range eggs			10	2015
Grain producer	Grain and buckwheat products			11	2015
Meat producer	Beef			12	2015
Vegetable producer	Potatoes, different kinds of vegetables			13	2015
Cheese producer	Traditional cheese, yoghurt			14	2015
Administrators
Founder of REKO				15	2015
Administrator in REKO Vaasa				16	2014
REKO Pirkanmaa (including several local nodes)
Consumers (gender)	Age	Occupation	Family status	Thread no.
A (female)	56	Researcher	Husband and one child	17	2016
B (male)	33	Sailor	Wife and two children	18	2016
C (female)	58	Unemployed	Husband, two grown-up children	19	2016
D (male)	26	Master builder	Girlfriend	20	2016
E (female)	42	Engineer	Husband	21	2016
F (female)	31	Engineer	Husband	22	2016
G (female)	44	Poet, multiple degrees	Husband, one grown-up child	23	2016
H (female)	55	Electrician	Husband	24	2016
I (female)	23	Student	Lives alone	25	2016
J (male)	44	Student	Wife and two children	26	2016
K (female)	34	Physiotherapist	Husband and two children	27	2016
Producers	Products
Meat producer	Beef (grain, eggs)			28	2016
Eggs and grain producer	Eggs, grain			29	2016
Meat producer	Beef			30	2016
Grain producer	Oat			31	2016
Vegetable producer	Berries, vegetables, fruits, potatoes			32	2016
Administrators
Administrator in several local nodes				33	2016
Administrator in several local nodes				34	2016

Appendix 1

Appendix 2. Interview questions

Consumers

• Background variables.

• Could you please describe how you make grocery shopping in your household?

• Please tell me, which issues are important to you when you shop for groceries?

• Tell me about your role and activity in the REKO network. When and how did you join the local ring and how would you describe your activities?

• Please tell me more closely, how and why you make grocery shopping via REKO? How would you describe your experiences so far?

Administrators

• Background variables.

• Could you please tell me about your role and activity in the REKO network? When and how did you hear about REKO and what was your role in developing the system? How would you describe your activities in REKO at the moment?

• Please tell me more closely how does the REKO network work? What are the principles of REKO? How do you select the producers to the ring? How would you describe the producers? How would you describe the REKO consumers? Why do you think they join REKO? Which are the challenges and benefits of REKO?

• What kind of future would you predict for REKO?

Producers

• Background variables.

• Please tell me about your farm and business: when have you started the farming, what do you produce, how much and where do you sell your produce?

• Tell me about your role and activity in the REKO network. When and how did you join the local ring and how would you describe your activities?

• Please describe your experience of REKO consumers. How would you describe your customers and your relationships with them?

• Please tell me more closely, how and why do you sell your produce via REKO? How would you describe your experiences so far? How would you compare REKO to other food distribution channels?