Collective knowledge brokering in social enterprises: developing an immersion-based model using Schein’s organisational cone

ABSTRACT

Knowledge acquisition and integration across boundaries is essential for developing innovative solutions in social enterprises (SEs). The vital role of knowledge brokers in this process has been sparsely discussed. This study examined knowledge brokering roles for effective solution development in SEs. We qualitatively analysed four Indian SEs using analytical techniques of grounded theory and the Gioia method. Brokering emerged as a collective process involving multiple agents who spanned for, translated and integrated two knowledge sets: expert and contextual. Correspondingly, two brokering roles emerged: boundary spanners and translators, which were further subdivided into primary and secondary. SEs displayed varying brokering configurations, which we explain by developing one of the first models of collective knowledge brokering using Schein’s organisational cone (1971) and immersion. SEs can employ our model to identify and groom agents into specific brokering roles, and develop beneficial brokering configurations, for enhanced effectiveness.

KEYWORDS:

• Collective knowledge brokering
• brokering roles
• immersion
• social enterprise
• middle managers
• Schein’s organisational cone

1. Introduction

Social enterprises (SEs) combine social objectives with business and entrepreneurial logics to develop innovative community-specific solutions to social problems (Dacin et al., 2011; Dees, 1998). Knowledge is critical for effective solution development in SEs (Maalaoui et al., 2020; Malsch & Guieu, 2019). Unlike commercial enterprises that protect and exploit knowledge for competitive advantage, SEs share knowledge more openly for enhanced value creation and social impact (Santos, 2012; Uygur & Marcoux, 2013). They also partner extensively across their boundaries for knowledge access and learning (Ostertag et al., 2021), employing in-person networks to share and manage knowledge (Bloice & Burnett, 2016; Buonomo et al., 2020; Ostertag et al., 2021). While commercial enterprises strategically develop efficient knowledge management systems, SEs use more informal knowledge sharing mechanisms (Granados et al., 2017). This lack of a structured and strategic approach to managing knowledge can undermine their objectives significantly (Buonomo et al., 2020) through knowledge losses. Yet research on how they acquire, interpret and utilise knowledge is rather limited (Granados et al., 2017; Maalaoui et al., 2020; Weerakoon et al., 2019).

Most research on managing knowledge is situated in large profit-oriented commercial enterprises that have clear strategic intent, organisational infrastructure and processes for knowledge management (Granados et al., 2017). However, these systems are different in smaller, lesser developed organisations, with differing missions and belonging to other sectors like SEs (Granados, 2015). First, SEs are usually small organisations with limited resources (Granados et al., 2017), so managing knowledge flows in them is likely to be more process and person driven (Bloice et al., 2016; Buonomo et al., 2020) rather than structure and infrastructure driven. Second, they acquire and utilise knowledge by actively interacting with stakeholders and knowledge resources often outside organisational boundaries (Kim et al., 2018; Maalaoui et al., 2020; Picciotti, 2017; van Twuijver et al., 2020). Consequently, they have more diffused and open organisational boundaries (Santos, 2012; Tandon & Nair, 2020). Third, SEs work closely with their target community (Mair & Martí, 2006) and their solutions are, and need to be, customised to community context and needs (Dees, 1998). These differences need to be incorporated to develop models for effectively managing knowledge in SEs (Granados, 2015; Granados et al., 2017).

In this context, knowledge brokers become vital for capturing, translating, and integrating different knowledge sets within and across diffused boundaries. Brokers bridge knowledge dispersed in disconnected networks, enable its sharing and recombination for learning and innovation (Hargadon, 2002; Jyrämä & Äyväri, 2007), and its interpretation and transformation into final outcomes (Obstfeld et al., 2014; Randhawa et al., 2017; Verona et al., 2006). Current SE literature does not explore the role of knowledge brokers; nor does it explain how individuals and processes can be structured to manage knowledge flows (Buonomo et al., 2020) within and across SE boundaries. We address this gap by examining the structures of knowledge brokering through a qualitative study of four Indian SEs.

We found that knowledge brokering in SEs was collective in nature, involving multiple types of knowledge that were spanned for, translated and integrated. Consequently, brokers positioned at different structural locations and variously immersed in the SE and target community contexts emerged. We explain these variations through Schein’s organisational cone (1971) and immersion. Using Schein’s cone, we mapped agent locations to the organisational hierarchy, and noted that just “any” agent at a given hierarchy level would not become effective broker by default – they needed to fulfil certain immersion requirements. Thus, we develop one of the first models of collective knowledge brokering, which can be employed by SEs to identify, appoint and develop effective brokers.

2. Literature review

2.1. Knowledge brokering

Knowledge brokering involves movement and interpretation of knowledge across organisational boundaries (Wenger, 1998, 2000), and its transformation into actionable knowledge (Patriotta et al., 2013). Brokers are individuals or entities that effect this process (Hargadon, 2002). Brokers belonging to and located inside the parent organisation/group do the following: identify and transmit information inside the organisation (Aldrich & Herker, 1977; Ancona & Caldwell, 1988; Tushman & Scanlan, 1981), represent the organisation externally (Aldrich & Herker, 1977; Ancona & Caldwell, 1988; Gould & Fernandez, 1989), regulate inflow of information (Ancona & Caldwell, 1988) and outsiders (Gould & Fernandez, 1989), and prevent information outflow (Ancona & Caldwell, 1988). Brokers also operate in the overlapping space between two groups while located inside one group or outside of both groups (Brown & Duguid, 1998; Gould & Fernandez, 1989). Recognising locational differences of brokers at the organisational boundary, Wenger (2000) identified four brokering functions: boundary spanners, roamers, outposts, and pairs. Boundary spanners locate themselves at one boundary and move knowledge across it, roamers move knowledge across multiple boundaries, outposts go beyond organisational boundaries and bring in information, and brokering pairs operate across boundaries of two interacting groups (Wenger, 2000).

These works assumed brokers as connecting hitherto disconnected groups. Brokers also mediate knowledge within organisations/groups (Obstfeld et al., 2014) like a “coordinator” within the group (Gould & Fernandez, 1989) and an “itinerant broker”, an outsider who mediates knowledge between group members (Gould & Fernandez, 1989). Shi et al. (2009) identify eight brokering functions when middle managers mediate knowledge between top and lower-level managers, their peers and top/lower-level managers, and within their peer group.

Further, varied nomenclatures accompany different brokering functions in literature. Haas and Claude Paraponaris Dr Martine Sigal (2015) identifies the most common terms as: boundary spanners, gatekeepers and knowledge brokers. Boundary spanners belong to one unit and link it to other units within or between organisations, facilitating information exchange, knowledge transfer, representation or resource access. Gatekeepers undertake environmental monitoring, regulate inward information flow, and transfer and diffuse knowledge inside. Knowledge brokers lie outside two groups and mediate knowledge between them. Overall, these studies indicate that the structural locations and affiliation of brokers impact their ability to broker knowledge.

Another literature thread discusses how brokers process knowledge. They may convey information across boundaries to create awareness between disconnected parties, transfer best practices, support knowledge sharing by drawing analogies, and enable knowledge recombination by synthesising information from different social worlds (Burt, 2004). Brokers may also utilise their control over knowledge for personal interests. Obstfeld et al. (2014) identified three such roles: conduit, tertius gaudens (who enjoys), and tertius iungens (who joins). The conduit transfers information/knowledge between two parties, tertius gaudens plays two parties against each other or keeps them separate for personal benefits, and tertius iungens facilitates ties between two parties and may be actively involved with them.

To summarise, brokers mediate knowledge within an entity, and with the outside world. They may be many and variously positioned with respect to the organisation/group, and process knowledge differently. Consequently, brokering literature has multiple overlapping nomenclatures and brokering functions (Haas & Claude Paraponaris Dr Martine Sigal, 2015). In this study, we examine which brokering roles emerge in the context of social enterprises.

2.2. Importance of knowledge brokering in social enterprises

Knowledge is vital for the accomplishment of SE objectives (Campos-Climent & Sanchis-Palacio, 2017; Malsch & Guieu, 2019). It promotes collaboration, innovation (Maalaoui et al., 2020) and performance (Do Adro et al., 2022) leading to higher social value creation and impact. Yet, SEs’ lack of a strategic approach, and processes and systems to support knowledge acquisition, sharing and utilisation (Buonomo et al., 2020), can limit their effectiveness greatly. There is inadequate understanding of the strategic, organisational and structural aspects of knowledge management in SEs (Buonomo et al., 2020; Granados et al., 2017; Renshaw & Krishnaswamy, 2009; Weerakoon et al., 2019).

SEs engage with and employ knowledge differently from commercial enterprises. Commercial enterprises manage knowledge strategically, including practices, processes and infrastructure for knowledge creation, sharing and institutionalisation (Begoña Lloria, 2008). They utilise and protect knowledge for competitive advantage (Uygur & Marcoux, 2013). SEs, on the other hand, have a social value creation mindset that promotes open knowledge sharing across their network (Bloice et al., 2016; Uygur & Marcoux, 2013). Most of this knowledge sharing is informal in nature (Bloice et al., 2016; Granados et al., 2017), and person-driven as SEs lack advanced technological and strategic knowledge management systems (Bloice et al., 2016; Buonomo et al., 2020; Weerakoon et al., 2019).

Knowledge sets needed for solution development are often located outside SE boundaries (Littieri et al., 2004), necessitating active cross-boundary interactions (Campos-Climent & Sanchis-Palacio, 2017; Hervieux & Turcotte, 2010) to acquire and utilise knowledge (Granados et al., 2017). Brokers, hence become pivotal for knowledge identification, acquisition (Kim et al., 2018), and its interpretation and integration into solutions (Jyrämä & Äyväri, 2007; Maaninen-Olsson et al., 2008). For successful brokering, brokers need to have an understanding of the contexts they engage with and mediate, and the ability to identify, transfer, translate and recombine relevant knowledge sets (Dobbins et al., 2009); alongside being able to gain access to knowledge.

In SEs, brokering roles are generally assumed voluntarily, even reflexively and are not formally designated. Such brokers may make sincere efforts, but they may not be effective due to the lack of relevant knowledge about the contexts they mediate, and about the strategic purpose and positioning of knowledge needed and available. SEs, therefore, need to purposefully identify and develop the right agents as brokers. SE literature does not provide adequate directions for this. Addressing this gap, we asked the following research question: How is knowledge brokered in Social Enterprises? Which brokering roles emerge and become significant for effective solution development?

3. Method

Data was collected from four Indian SEs identified from the Ashoka Foundation database (www.ashoka.org) and anonymised as: Entrepreneurship Development Network (EDN), Education Foundation (EF), Society for Social Action (SSA) and Vision Foundation (VF). Since knowledge sharing and learning interactions are central to innovation in SEs (Maalaoui et al., 2020), we chose SEs based on the innovativeness of their projects. This was also confirmed in initial interactions with the SE heads to ensure that the chosen projects did involve learning interactions. Each SE had three broad management levels: top management (including CEO(s)), middle managers heading regional offices and departments, and operational staff. We chose individual offices in each SE that were comparable in size. Table 1 provides details of the SEs, projects studied, participants and data collection sources.

Table 1. Data collection details.

Social Enterprise	Sector	Objective	Project studied	Office studied and Employees interviewed	Innovative solutions developed involving brokering
EDN (Established: 1995)	Economic development	Economic development of underdeveloped regions	Microfinance and self-help groups (SHGs) program	One regional office 9 interviews; 2 group discussions including 25 employees Participants: CEO (1); Head office support staff (3); Regional office head and middle managers (department heads) (5); Operational staff (25); Target community (field visit interactions)	SHG processes Microfinance loan, Other community solutions
EF (Established: 2008)	Education	Improving primary education through developing headmasters of municipal schools	Headmaster capability development program	One regional office 6 interviews; 2 group discussions including 6 long term interns (operational staff) Participants: Middle managers (department heads) (6); Operational staff (6)	Leadership development program
SSA (Established: 1997)	Education (part of larger community development)	Bringing children of migrant workers into the mainstream education system	Innovative education system	Head office, and regional office for the district 5 interviews; 1 group discussion including 8 employees Participants: CEOs (2); Middle managers (department heads) (3); Operational staff (8); Target community (field visit interactions)	Progressive education program
VF (Established: 2002)	Health	Social and economic inclusion of the visually impaired	All activities to support the visually impaired	Head office, the only functional office 7 interviews Participants: CEO (1); Middle managers (department heads) (5); Long term intern (1); Operational staff (informal interactions)	Solutions for supporting visually impaired

To identify projects, we consulted the CEOs, and in one case the regional head. The first author (henceforth “researcher”) immersed herself for 2 weeks in each SE, conducted semi-structured interviews and group discussions, undertook field visits, made daily observations, participated peripherally in activities, and engaged in informal conversations with the employees and target community. She noted her observations and experiences in a field diary. This primary data was cross validated through focussed conversations across multiple levels, mostly with middle managers, and wherever possible, with the CEOs. It was further supplemented through secondary materials from SE websites and annual/periodic reports, reporting formats and other publications/documents shared by the SE.

Semi-structured interviews were aimed at eliciting learning episodes, which are events, actions and experiences resulting in tangible or intangible learning outcomes (Sole & Edmondson, 2002). Interviews started with exploring the participants’ perspective of SE’s objectives, their individual roles, followed by in-depth discussion of specific projects. The researcher encouraged and probed them to share how the projects came about, their role(s) and experiences, instances of new activities, processes, solutions being developed, as well as stories of success and failures experienced. A few group discussions were also conducted with the operational staff. All interviews and group discussions were recorded and transcribed verbatim and lasted between 25–85 minutes, averaging at 43 minutes.

Both authors analysed the data. First, we read the text and classified it into learning episodes (Sole & Edmondson, 2002), identifying descriptions that led to changes in mental models and development of new processes/practices/products/services. Next, we analysed the episodes using the constant comparative method (Glaser & Strauss, 1967). We compared each episode with the next to develop categories and compared emerging categories within and across SEs. We constantly questioned and reviewed categories to modify, broaden and streamline them. Next, following Gioia et al. (2013), we clubbed these categories into higher level themes to develop a data structure (Figure 1). During this process we constantly iterated between data and theory (Eisenhardt, 1989) and identified collective knowledge brokering as the central theme.¹

Figure 1. Data structure.

We took several steps to establish the trustworthiness of our analysis. Both authors continuously reviewed emerging themes critically from multiple theoretical lenses. Differences in coding were reconciled through discussions. We also presented the analysis to three external experts for their critical assessment, re-coded the data and presented it to them again until a common understanding of the themes was reached.

4. Analysis and results

Two types of knowledge – expert and contextual, were acquired by SEs, interpreted and integrated for solution development. Expert knowledge is domain specific knowhow acquired through education and professional experience (Tandon & Nair, 2020), like microfinance or primary education pedagogy. Contextual knowledge is comprehension about the socio-economic, political, cultural, and psychological context of the target community (Tandon & Nair, 2020). Their integration into utilisable knowledge was driven by strategic knowledge possessed by specific agents in the SE. Strategic knowledge refers to knowledge about the organisation’s current and future goals, which enables alignment of current activities with SE objectives. We identified brokering roles and mechanisms enabling the movement and integration of expert and contextual knowledge. Two brokering roles emerged: boundary spanners (subdivided into primary and secondary in two SEs), who scouted for and moved external knowledge sets inside the SE, and translators, who translated, i.e., interpreted and integrated them. These roles were enacted by different agents positioned at different structural locations in the SEs. Thus, brokering emerged as a collective process enacted by multiple brokers. We describe the brokering configurations for each SE below.

4.1. EDN

EDN supported microenterprise development in underdeveloped districts in the Indian States of Uttar Pradesh and Bihar. It provided microfinance services through self-help groups (SHGs) and undertook health awareness, education and agriculture interventions.

The CEO of EDN had previous experience in creating and managing SHGs and started by forming them in the target community. SHGs are informal groups who come together for an identified purpose and eligible for microfinance. EDN created primarily women SHGs to promote monetary savings and support microenterprise development. For starting other programmes such as microfinance, health and agriculture, where the CEO did not have expertise, he scouted for and involved external experts, thus assuming primary boundary spanning role. For instance, he arranged training programmes with a partner organisation to train middle managers on concepts and processes of microfinance and public health, who then trained the operational staff. Once trained, the middle managers also scouted for expert knowledge but not as a primary role. For example, the agriculture manager described how he learnt about the System of Wheat Intensification (SWI) method of growing wheat and worked for its adoption by farmers:

I was working in another district. Sir (the regional office manager) asked me to work here [current region] and help farmers improve production using the SWI method. I am working on [identifying new agricultural techniques to improve production of] wheat, rice, mentha, peppermint, medicinal plants etc. I meet farmers to convince and help them adopt the technology. (Expert knowledge, Secondary boundary spanner)

This indicated the emergence of two boundary spanning roles differing in responsibility for spanning: primary boundary spanning by the CEO, and secondary boundary spanning by middle managers.

EDN innovated on the use of SHG from an instrument of saving, to an instrument for complete social intervention. The SHG helped EDN to access all members of a household and develop interpersonal relationships resulting in deep understanding of the community context. EDN hired target community members as operational staff to root itself strongly in the community. These employees possessed in-depth knowledge about the community and executed EDN’s programmes on the field. They worked closely with middle managers and were knowledgeable about EDN’s objectives. This enabled them to effectively span the target community boundary, identify relevant contextual knowledge, and transfer it within the SE to middle managers, thus emerging as primary boundary spanners for contextual knowledge. Middle managers worked closely with the operational staff through regular interactions and field visits, thus assuming a secondary boundary spanning role. Thus, EDN was able to capture and utilise contextual knowledge to develop community specific processes, like in microfinance:

Microfinance manager: When we started giving loans [to SHGs], we [middle managers and operational staff] saw that initially the loan is used for [personal] consumption. Then, when the SHG has enough money, they take loans for gaining assets: quilts, pillows, utensils, etc. When more money gets accumulated, they think about income generation activities. We tell our [operational staff] to continuously look for these things: whether children are going to school, are their living and health conditions improving, etc. (Contextual knowledge, primary boundary spanner: operational staff). If not, there will be no progress [the purpose of giving loans will not be met]. We [middle managers] ask about this and we stop/delay their loans. [if living conditions are not improving]. (Secondary boundary spanner: Middle manager)

Once expert and contextual knowledge came inside the organisation, they were translated by interpreting and integrating them and aligning them to the strategic objectives of EDN. As they had in-depth understanding of EDN’s strategic orientation, and possessed both expert and contextual knowledge, middle managers emerged as translators of knowledge. For instance, the regional office head, explained how they learnt to go beyond theoretical knowledge to create successful women SHGs. Decision making by women in rural Indian families was not individual driven, but family driven. Thus, to participate in SHGs, their families also needed to be on board:

In early days we faced many roadblocks in SHG formation … there were many books on how to form SHGs, there were NABARD norms. But when we started on the ground, we realised the difficulties. If we are making SHGs but not taking the family into consensus, then that SHG would not be successful, it will only be on paper. We realised that we were working in theoretical way but leaving out the family part. We started explaining the need for SHGs to the families [instead of only women], then they started taking interest. (Integration of expert and contextual knowledge, Translator: Middle Manager)

Similarly, middle manager microfinance spoke about developing norms for Joint Liability Groups (another social group created purely for loan disbursement) by understanding the social-familial setup:

We saw that if two people from the same family are taking loans, that created problems. If they fight, then one brother won’t be able to tell the other to repay the loan. Or they may come together and refuse to repay. So, we created a norm that we won’t give loans to more than one person of the same family in JLG. (Integration of expert and contextual knowledge, Translator: Middle Manager)

Thus, in EDN, boundary spanning for both knowledge sets was distributed into primary and secondary roles. For expert knowledge, CEO was the primary boundary spanner while middle managers acted as secondary role takers. Similarly, operational staff acted as primary boundary spanners for contextual knowledge with middle managers taking a secondary role. Finally, middle managers emerged as translators of these knowledge sets resulting in the development of context-specific solutions.

4.2. EF

Operating in the Indian States of New Delhi, Gujarat, Maharashtra and Rajasthan, EF worked towards improving quality of primary education in government schools. It ran customised programmes for developing leadership capabilities in headmasters, which in turn impacted the school’s quality of education. EF created a competency framework for headmasters which was implemented through a leadership programme.

EF’s CEO had prior experience in education sector but needed expert inputs to develop its programme. He partnered with external experts like consulting firms and faculty from a reputed management institute to develop a leadership competency model and the programme structure. He also connected them with EF’s middle managers; thus emerging as the primary boundary spanner. Middle managers acquired expert knowledge on primary education pedagogies through mutual engagement and co-creation with the experts; also assuming a secondary boundary spanning role as they scouted for programme related expert knowledge. A middle manager explained:

We work with our partners to create programs specific for our requirements. We have good partnerships with BDH and JG [external expert organizations]. We have learnt a lot from them. [for classroom learning and maths pedagogy respectively]. (Expert knowledge, secondary boundary spanner: Middle managers)

To effectively implement their programme, EF needed in-depth understanding of the government school setup and their target community. So, the operational staff (interns joining EF for a two-year fellowship), were immersed in the target community for two months, followed by regular field-based interactions. Unlike EDN, the operational staff were not target community members, and needed specific immersion processes to internalise the community context. For a month, they taught at the schools and engaged with the children’s families (living in underdeveloped communities). This was followed by a month of “slum immersion” during which they lived in the community, and navigated their way through it, acquiring in-depth understanding of the target community context. A middle manager explained:

Slum immersion is a time [when operational staff] get an idea of teachers’ and parents’ psyche. There is lot of bias that parents are drunkards, they are not interested in getting their children educated, [but want] their children to work somewhere. So are they able to get into the slum and understand what motivates parents to behave the way they are? Is it cultural? Is it sociological, what is their need?. (Contextual knowledge, immersion)

After immersion, operational staff visited the schools daily and the community regularly. These processes enabled them to identify relevant contextual knowledge and transfer it to the middle managers, thus emerging as the primary boundary spanners. Middle managers also spanned boundaries occasionally (secondary boundary spanners) through field visits. A middle manager explained the boundary spanning role of operational staff:

The [operational staff] implement the program in the field and then they debrief what happened with the Headmaster with us. This debrief happens mostly on the same day they visit the field, otherwise on alternate days. (Primary boundary spanner: operational staff, contextual knowledge)

The debrief was an intensive process of action-reflection where each action and observation of the day was reflected upon, deconstructed and discussed to identify issues, and to plan the next course of action.

Middle managers’ role design necessitated them to supervise operational staff and head departments linked to the leadership programme. Consequently, all were deeply integrated into field operations and the leadership competency framework, thus possessing both expert and contextual knowledge. They met formally every week and informally anytime to reflect on emerging issues. They had weekly and need-based daily action-reflection sessions with the operational staff to understand field issues and to place them in the programme’s perspective. This enabled them to translate knowledge by combining expert and contextual knowledge, to develop customised operational plans and review/modify their competency framework. For instance, they realised that their linear competency development approach was not working on the field; instead, multiple competencies were being impacted together. So, they planned to change the model to align with the field experience. Similarly, operational staff explained how customised plans were developed for each Headmaster:

Before the academic session [year] starts, we have a 3–4 days [Kick Start] workshop with all headmasters (conducted by middle managers, sometimes with external experts). In that, they are exposed to certain content [technical inputs on competencies, pedagogical tools, school improvement interventions], so are we. After the workshop, we sit and plan with them [create an action plan], aligning the workshop content with their need and of the school.

This plan was reviewed by middle managers and other operational staff, then discussed with the headmaster again. This process was iterated to create a final action plan. This led to the integration of expert and contextual knowledge, where middle managers led the translation process and emerged as translators.

Similarly, a middle manager described how she and internal subject matter experts (also at middle management level) translated knowledge sets to develop the “Kick Start” workshop:

When we design the workshop, I from operations team and a curriculum person [subject matter expert], sit together. I tell her that this is happening in the field, this is the requirement … then she comes up with her expertise and we integrate it. (Translation, Translator: Middle manager)

The process of action-reflection, mutual engagement and co-creation was evident in all interactions between the middle managers, operational staff and headmasters. It enabled alignment and contextualisation of the leadership programme to each headmaster and school.

Thus, in EF, the CEO emerged as the primary boundary spanner for expert knowledge, followed by middle managers as secondary spanners. Operational staff emerged as primary boundary spanners for contextual knowledge, with middle managers performing a secondary role. Translation was undertaken dominantly by middle managers.

4.3. Knowledge brokering in EDN and EF: the emergence of a model

The above analyses showed knowledge brokering as a collective process, involving multiple brokers spanning for and translating different knowledge sets: expert and contextual. Boundary spanning was split into primary and secondary roles and assumed by operational staff or middle managers. The translation role was assumed by the middle managers. We examined these configurations using the concept of immersion (which emerged as an in vivo code from EF), and Schein’s (1971) model of organisational socialisation. Schein (1971) described the organisation as a cone. Employees were located centrally or at peripheries of the cone, both vertically and horizontally. He mapped socialisation and career movements of employees vertically (up the hierarchy); radially (to more or less central roles); and circumferentially (between different units or departments). His model highlighted the importance of structural locations and hinted towards variations in immersion in the organisation. Immersion is a process by which an individual experiences a social group as a native, without losing his/her identity (Glaser & Strauss, 1967). This group could be a community, an organisation or a team. The immersed individual experiences the socio-cultural-psychological-behavioural fabric of the group and develop tacit knowledge about it.

We mapped the structural locations of brokers using Schein’s model, and examined their immersion levels in the SE and target community, to infer why/how these brokering roles were effective in EDN and EF. Figure 2 illustrates the location and immersion of brokers in the two SEs. The flat organisational cone depicts the SEs’ relatively flat organisational structure. The target community is added as another layer with dotted lines and positioned closer to the SE to indicate its active involvement in SE activities. Solid black dots represent agents deeply immersed in the SE: the middle managers and top management team (CEO) who were centrally immersed along all three dimensions of Schein. The hollow dots represent the operational staff. They were immersed in two contexts: the target community, and the SE. They were vertically and radially distanced from the SE’s centre, thus indicating low immersion in the SE, but were deeply immersed in the target community.

Figure 2. Collective knowledge brokering (EDN & EF).

Boundary spanning for expert knowledge required strategic knowledge of the SE goals. To be effective, the primary boundary spanner needed to be centrally immersed in the SE, vertically and radially, and was therefore, the CEO in both the SEs. Middle managers who were sufficiently immersed in the SE vertically and radially, had strategic knowledge, and acted as secondary boundary spanners. Boundary spanning for contextual knowledge required deep immersion in the operational context, the target community, and broad awareness of the knowledge requirements of the SE. The operational staff, located at the vertical periphery of the SEs and deeply immersed in the target community played this role effectively. By closely working with the operational staff and the target community, thus immersing sufficiently in it, the middle managers assumed a secondary boundary spanning role. Knowledge translation required the broker to possess both knowledge sets (expert and contextual) and the strategic knowledge to interpret, integrate and align them with the SE’s goals. Middle managers, being secondary boundary spanners for both types of knowledge, and being centrally and vertically immersed in the SE, enacted this role effectively.

We refined and substantiated this model further, as we employed it to explain the variations in boundary spanning and translation roles observed in the next two SEs.

4.4. SSA

SSA worked with migrant labourers in stone quarries in the Indian State, Maharashtra, to improve their living conditions through multiple interventions. We studied one of their central programmes: regularisation of education for the children.

The CEOs (founder and co-founder) had educational degrees in social work but no knowledge of the context of stone quarry workers. They started with daily visits to the community to experience and understand their issues. They identified education as a central need and approach to initiate larger change in the community. Since parents were migrant labourers, the children frequently dropped out of schools, could not continue regular education, and therefore needed a different kind of educational programme. To develop this, the CEOs connected with external expert institutions, studied their curricula, and underwent training in nonformal teaching, thus emerging as primary boundary spanners for expert knowledge. They disseminated it to the middle managers and operational staff through internal trainings and interactions. The CEO explained:

We both [CEOs] have a background of social activism, and we are qualified in the same line [Masters in Social Work]… we also tried to get external inputs to learn. AIE [external expert organization] – we underwent the first training programs in non-formal teaching through them. (Expert knowledge, boundary spanner: CEOs)

The CEOs also connected their middle managers and operational staff with pedagogical experts when required, such as for teaching maths and science. As the CEOs spent considerable time initially in the target community, they gained contextual knowledge which enabled them to develop the innovative education programme. Even after programme operationalisation, they continued regular field interactions with the community; thus spanning the target community boundary consistently. Simultaneously, the middle managers and operational staff also spanned target community boundaries (both positions employed target community members). Middle managers had a coordination role rather than a strategic role, worked alongside the operational staff, supervised them and spanned boundaries along with them. They shared the knowledge gathered with the CEOs in regular meetings. Thus, in SSA, primary and secondary boundary spanning roles did not seem to be clearly differentiated. Instead, they were equally distributed across all the three levels.

The CEOs, by virtue of possessing expert, strategic and contextual knowledge, emerged as translators. This was reflected in their innovative education programme. They identified that the children of migrant labourers dropped out of schools due to the migratory nature of their parents’ job. Also, as children grew old, they would feel ashamed to attend grades which were meant for younger children and would drop out. The CEOs combined this information with their expert knowledge and developed the “progressive education program” that enabled students to complete multiple grades in one year until their age matched their grade. Then they were facilitated to get enrolled in government schools. The CEOs explained:

… we have a training package [for teachers]. With experience [we also] developed a teaching methodology which is multi-class teaching using the Paulo-Ferrerian nonformal method of teaching… it is not theoretical but contextualized. (Integration of expert and contextual knowledge, translator: CEOs)

Thus, in SSA, a different configuration of knowledge brokering roles emerged. CEOs were the sole boundary spanners of expert knowledge. Contextual knowledge was spanned by all three levels, with no emergence of primary or secondary spanning roles, and translation was undertaken by the CEOs. Applying our model to SSA, we observed that since the middle managers were not immersed sufficiently vertically or centrally into the SE context, they could not act as effective translators. Instead, the translation role was totally taken up by the CEOs who were highly immersed in the SE along all of Schein’s dimensions, and maintaining significant immersion in the target community. Since immersion in the target community was maintained across all levels, so was boundary spanning for contextual knowledge, and primary and secondary roles were not differentiated (Figure 3).

Figure 3. Collective knowledge brokering in SSA.

4.5. VF

VF acted as a knowledge resource for the visually impaired to facilitate their social and economic inclusion in the society. It operated through a helpdesk, a website, an advocacy wing, a research wing, and a radio programme.

In contrast to other SEs, limited learning interactions were observed in VF. Interactions with external experts focused on information sharing rather than learning. Boundary spanning for expert knowledge was distributed across different departments as information gathering and dissemination was central to the SE. In addition, VF had developed a formal boundary spanning process for expert knowledge. The CEO and middle managers identified similar organisations and sent employees (irrespective of their position in the SE) to scan for, identify and bring back relevant knowledge. They would then discuss their observations with all employees in a meeting. However, not much learning seemed to be derived from this activity. The researcher observed one such meeting:

CEO:

Which features [of the visited organisation] did you notice for ease of access for visually disabled?

Operational staff:

Nothing specific

Middle Manager-1:

Any detail they shared under advocacy?

Operational staff:

No; I didn’t ask also

Middle Manager-2:

How can our channels benefit from them?

Operational staff:

For helpdesk [team]- we receive calls from multi-disabled, we can connect with them. For website [team]- we can share the resource centre [no further details described]

CEO:

What about their approach? Thinking, philosophy, any ideas that we could borrow … ?

Operational staff:

I wasn’t prepared for this … I don’t know

(Expert knowledge, boundary spanner: operational staff, ineffective boundary spanning).

The questions asked by middle managers required strategic knowledge, which the operational staff did not possess, as they were not immersed centrally into the SE context.

When we examined contextual knowledge, a similar deficiency was observed even though VF employed visually disabled people (the target community) at all levels (including the CEO). Clear boundary spanning and translation roles were not observed. The target community was diverse and contact between VF and the community was mediated by a telephone line. There was no consistent immersion and field-based interaction with the community. This resulted in generic solutions without specific contextualisation as per community needs, as VF was not able to successfully interpret their expert knowledge in the target community context. In those few instances where VF came in direct contact with the target community, middle managers and sometimes the CEO, spanned the boundary and acted as translators, and effective solutions were developed. This was evident in the following unique instance of effective brokering leading to changes in a workshop content:

CEO-VF: We [CEO and a middle manager] were recently doing workshops in Odisha [an Indian state] …what we were doing we felt it was important for them, [but] they were in a situation where living life with disability was a huge effort, therefore we needed to mentally prepare them with baby steps rather than showing them the sky. So, we had to draw back and come forward with little steps for them. (Integration of expert and contextual knowledge, Translator: CEO and middle manager)

Need for understanding the target community context was also felt by other middle managers who had undertaken field visits to conduct information sharing workshops. For instance, the advocacy team opined:

Unless we have been into the field talking to people, we wouldn’t actually understand that the[ir] need is something else that has to be taken care at its first level.

This lack of immersion in the target community resulted in ineffective brokering. Although boundary spanning was happening for expert knowledge, there was no clarity on its objective since contextual knowledge was lacking. Thus, in VF, brokering for expert and contextual knowledge was ineffective.

Applying our model to VF (Figure 4), we see that the distance between the operational staff, middle managers and target community was large due to no immersion in the target community context. This reduced the ability of operational staff to act as effective boundary spanners and the middle managers to translate knowledge. Further, operational staff not being centrally immersed in the SE, were also not the best choice for spanning expert knowledge boundaries as revealed in their effort failures.

Figure 4. Minimal knowledge brokering in VF.

5. Discussion

SEs are knowledge intensive organisations (Weerakoon et al., 2019), with differing objectives and confronting unique challenges for knowledge sharing, compared to commercial firms (Bloice et al., 2016). Their effectiveness and capacity to innovate depends on their ability to strategically share and manage knowledge (Buonomo et al., 2020; Maalaoui et al., 2020); yet this literature is largely underdeveloped (Langmann et al., 2021). SEs share knowledge across boundaries to create larger social value (Uygur & Marcoux, 2013), and capture knowledge resources often located outside their boundaries to develop innovative context-specific solutions (Campos-Climent & Sanchis-Palacio, 2017; Tandon & Nair, 2020). Most SEs actively acquire knowledge, but lack the ways to effectively interpret, convert or apply it (Granados et al., 2017) due to underdeveloped microstructures to process knowledge. This study contributes by detailing some of these specific structures of knowledge brokering that enable effective knowledge interpretation and use.

First, we show that knowledge brokering in SEs is collective in nature and involves multiple brokers (boundary spanners and translators), variously positioned and immersed vis-à-vis the SE and the target community. These individuals do not always belong to the same team, are located across hierarchy levels, and immersed diversely in multiple related contexts (here, SE and target community). This finding augments emerging literature on the collective nature of brokering (Wye et al., 2020), which only a few studies have examined, such as in teams (Harvey et al., 2014; Joshi et al., 2009; Kislov, 2018; Marrone, 2010), and healthcare (Bishop & Waring, 2017; Currie & White, 2012; Currie et al., 2015).

Second, we develop one of the first models of collective knowledge brokering in SEs using Schein’s organisational cone (1971) and integrating it with the concept of immersion, which provides conditions under which specific actors become effective brokers, and the brokering tasks they undertake. As organisations move beyond strong boundary conceptions to more diffused ones, immersion becomes an important indicator of the knowledge held by individuals irrespective of the nature of their engagement with the organisation. Combined with Schein’s cone to identify structural location vis-à-vis the organisation, it provides a clear picture of which individuals can be effective brokers for different types of knowledge.

Third, we contribute to emerging research on knowledge sharing and management in the social enterprises sector. Developing effective structures to utilise knowledge is as essential in SEs as in commercial enterprises. Considering the dispersed nature of knowledge resources across boundaries, limited fund/resource availability (Granados et al., 2017), and a person-driven approach to knowledge sharing (Buonomo et al., 2020; Ostertag et al., 2021) in SEs, our study proposes adoption of a collective and an immersion driven approach by them to create structures that capture and utilise knowledge effectively.

5.1. Collective knowledge brokering in social enterprises: an immersion-based approach

We highlight immersion as a critical factor for effective brokering. Immersion is used in language studies to teach a second language (e.g.: Genesee, 1985; Kearney, 2010). It is also a central approach in ethnographic research (Hammersley & Atkinson, 2019). In management studies, it is discussed sporadically in expatriation research (e.g.: Bird & Dunbar, 1991), and organisational culture change (e.g.: Wilkinson et al., 1996). It has not been spotlighted for its role in product/service/solution development, although the importance of tacit knowledge in situated contexts imply immersion. Similarly, SE literature recognises the importance of context embeddedness (Mair & Martí, 2006), and the role of stakeholders in innovating (e.g.: Datta & Gailey, 2012; Khavul & Bruton, 2013), implying immersion but not discussing it specifically. Our study brings this concept to the forefront for knowledge brokering.

In this study, immersion had two facets: immersion as a precondition for capturing contextual knowledge, and immersion in the organisational context (possessing clarity and knowledge of organisational strategic goals) as a precondition for the ability to translate knowledge fittingly for the organisation (Figure 5). Immersion in the target community enabled agents to experience its specific socio-cultural-economic-political context. This tacit knowledge formed a critical input to solution development. SEs used two modes of immersion: “community employment” and “lived experience” in the community. Effective learning occurred when these modes were followed up by regular field-based interactions and engagement with the target community ensuring sustained immersion. Effective immersion was not achieved by merely employing the target community; instead, lived experience was essential, thus indicating the importance of locational and interactional proximity of employees to the target community. Employees belonging to the target community, if distanced from the social context of operations, brought in only partial information based on their personal experience which differed from the actual community context, as seen in VF. Immersion was also a precondition for effective translation of expert and contextual knowledge. The ability of middle managers and top management to interpret and integrate them from the SEs’ perspective depended on their lived experience as employees located centrally in the organisation and in decision making roles. This enabled development of strategic knowledge and understanding about the SE context, which was critical for effective translation.

Figure 5. Immersion as a precondition for knowledge brokering.

Taking guidance from our study, SEs can develop modes of immersion in relevant communities to effectively span boundaries. Within the SE, they can identify critical translation roles and develop middle and senior managers into these roles by immersing them in critical projects and decision-making roles, thus enabling them to acquire relevant strategic knowledge and perspective.

5.2. Middle manager as knowledge broker

This study reinforces the significance of the middle manager as a critical interpreter of knowledge (e.g: Beck & Plowman, 2009; Sharma & Good, 2013), and extends it by identifying their specific brokering functions (Shi et al., 2009), thus contributing to Radaelli and Sitton-Kent’s (2016) call to examine the translation roles of middle managers. Like hybrid middle managers (Burgess & Currie, 2013), middle managers in our study performed three critical functions: downward movement of external (expert) knowledge inside the organisation, upward flow of knowledge from local context, and translating these two knowledge sets in the context of organisational goals. Further, we show that “any” middle manager cannot be an effective translator: only those immersed centrally in the organisational context, working closely with operational staff, and regularly immersed in the target community are likely to have this capability. Thus, we provide suggestions towards identifying, appointing and developing the “right” middle manager into brokering roles.

5.3. The role of ‘trust’ in brokering

As we reflected on our data, we observed that establishing trust with the target community was a vital enabler of brokering. SEs employed various mechanisms to build trust and signal authenticity of purpose which enabled them to gain access to contextual knowledge via immersion.

EDN used the following mechanisms to signal trust: displaying interest in the larger good of the community, actively involving the community through hiring, and personal relationship development. For instance, for giving microfinance loans, they checked if the person’s daughter(s) went to school. Their loan recovery norms signalled their disinterest in the loanee’s material belongings (they did not partake food/water from their household). EF had an intensive process of action-reflection and open team environment where hierarchies did not matter, where middle managers and operational staff routinely deconstructed their actions, behaviours, and experiences. This was also extended to the headmasters. The operational staff would observe staff meetings in schools, post which they would intensively deconstruct it with the headmaster to examine the events that occurred, the reaction(s), the outcome(s), and why and what could have been done differently. This was followed by joint action planning and course correction. This openness enabled development of trusting relationships between the headmasters and EF.

In SSA, the CEOs initially spent considerable time in the community and helped in daily chores to get children to their school, thus indicating their intent and authenticity. They were also involved in social activism for the target community. The CEOs’ deep relationship with the community was witnessed during the researcher’s visit, when the marriage ceremony of a community girl was conducted in their own house. The researcher also observed the community involving the CEOs in solving interpersonal and community issues. In VF, unlike these other SEs, such instances were not observed or narrated in the interviews.

Trust, hence, is a significant factor in facilitating knowledge exchange in SEs (Weerakoon et al., 2019). In consonance with Heinze et al. (2016), we also emphasise the importance of developing trust with the target community to thoroughly understand local needs for developing authentic context-specific solutions. In its absence, brokers would have limited or superficial access to contextual knowledge, thus constraining or even crippling effective brokering.

6. Conclusion

Our study contributes to SE literature on knowledge sharing and management, and the larger knowledge brokering literature, by developing one of the first models of collective knowledge brokering. Innovative solution development involves multiple types of knowledge, such as of market (Cillo, 2005), context (Brown & Duguid, 1991) and technical, harnessing each of which requires employment of different brokering roles (Kislov et al., 2016). Our model substantiates this need for identifying different brokers and roles for different knowledge sets. Brokering literature also depicts variations in the embeddedness of brokers vis-à-vis organisations: brokers can be located in overlapping boundaries between entities or inside an entity (e.g.: Brown & Duguid, 1998; Cillo, 2005; Shi et al., 2009). Our model aids in reconciling these variations through the concept of immersion.

Our model is of direct use to SEs and other similar organisations to identify and develop suitable agents as effective brokers. By examining the type of knowledge needed, the context in which it is embedded, and examining the immersion of various agents, brokers who are most likely to be effective can be identified and groomed into the role. By developing a network of brokers, organisations can enhance their ability to effectively capture and utilise knowledge.

We examined individual small offices in SEs with flat structures. Inferences from our study can therefore be extended to SEs and similar other organisations with small offices and flat hierarchies. With changes in organisational size and hierarchies, brokering roles and conditions for effectiveness are likely to vary. However, immersion is still likely to remain a critical factor in understanding brokering. Further studies to examine brokering roles in various organisational contexts using immersion as a lens is likely to enrich our understanding of collective brokering. Research on the role of trust in brokering is also needed. We did not examine factors like legitimacy of brokers (Levina & Vaast, 2005) and role of power differentials (Currie & White, 2012), which too can impact brokering effectiveness. Research incorporating these aspects can further broaden our understanding of brokering and related knowledge management microstructures in SEs for higher solution effectiveness and value creation.

Declaration

The researchers followed ethical guidelines of research during this study. Participants involved in this study were adults who were cognitively capable of making decisions. Informed consent was sought from participating organisations before starting the study. Informed consent was also taken from each participant before they were interviewed. All participants were informed about the purpose and academic usage of the data. The research followed the ICJME guidelines on protection of research participants, the Belmont Report and the Helsinki declaration revised in 2013 in terms of the maintaining the ethics of the research process, informed consent and confidentiality. Further, the findings were discussed at the organisational level and not at the individual level. It was not medical research, did not involve interventions of any kind and did not attempt to modify behaviour. Therefore, clauses from these guidelines related to medical and behavioural intervention research were not applicable for this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Notes

1. This study was part of a larger qualitative study on organisational learning in social enterprises from which multiple key themes emerged. In this paper, we discuss one: knowledge brokering, as the central theme.