The rise of CSs

In the past few decades, we have witnessed an increased scholarly and managerial attention to how spatial and material realities can affect how organizations operate and trigger changes in employees’ behaviors (Clegg & Kornberger, 2006). Although the first studies on organizational space focused on how the physical characteristics of the space where people ordinarily work affect behaviors (e.g., Davis et al., 2011; De Croon et al., 2005; Oldham & Rotchford, 1983), a more recent scholarly conversation has started to focus on the phenomenon of CSs (coworking, incubation spaces, social innovation centers, fab labs, cultural centers, technology parks, etc.). In the digital organization era, CSs respond to the needs for new organizational knowledge practices, increasing flexibility, and faster responses in the environment. It is interesting to notice that these spaces have been often referred to as “third places.” Accordingly, the CS allows workers to experience and reproduce organizational practices even if they are not part of an organization (Cnossen & Bencherki, 2019).

Till now, potential benefits of CSs have been mainly targeted to knowledge workers, gig workers, and micro start-ups. For knowledge workers and gig workers, these spaces may be an affordable workspace which may offer visibility (Merkel, 2019), legitimacy, and a sense of community (Garrett et al., 2017), thus contributing to avoiding the risks of isolation and job insecurity as well as to building a professional identity (Petriglieri & Obodaru, 2018). Similarly, CSs may satisfy the need for efficiency of start-ups and their desire to join a network of professional relationships which might sustain and accelerate their business ideas (Capdevila, 2015; Spinuzzi, 2012). The systematic hybridization between the internal competences and the external heterogeneous competencies provided by the different actors who work in the CS may sustain creativity, contributing to reducing inertia and avoiding the mere replication of past solutions (Gavetti & Levinthal, 2000).

Organizational workplaces as “third places”: The boundary paradox

It is noteworthy that recently CSs have also started to inspire companies to design more open internal workspaces with physical characteristics (e.g., architecture and layout) and functioning logics that recall “third places” (e.g., flexible use of space and time; calls for ideas or innovation projects). Accordingly, extant research on CSs distinguishes between external, commercial coworking spaces (Bouncken et al., 2016) and internal, corporate CSs, i.e., CSs within large corporations such as internal (open) coworking spaces, corporate fab labs, or corporate hybrid spaces (de Vaujany et al., 2019; Leclercq-Vandelannoitte & Isaac, 2016). However, to the best of our knowledge, extant studies have focused mainly on external CSs and the few empirical evidence on corporate CSs is mainly explorative. Nevertheless, extant literature recognizes that the use and the practices that construct these spaces can be very different (Bouncken et al., 2018; Jakonen et al., 2017). Thus, for instance, organizations may adopt the model of the CS to communicate aspects such as openness to external stakeholders (e.g., customers, supplier, and local community), social responsibility and non-territoriality with respect to internal conflicts between organizational functions (e.g., R&D, marketing, and finance) (Elsbach, 2003; Elsbach & Pratt, 2007; Leonard, 2013). Given the importance for organizations of managing their internal and external boundaries, we argue that such new trend entails an unexplored paradox interesting to study.

On the one hand, setting up CSs inside organizations aims at breaking down different types of boundaries, for instance, those between different organizational functions to stimulate the exchange of knowledge and co-creation, by making hierarchies flat, team collaborations more spontaneous, and individuals more willing to participate in team work (e.g., Capdevila, 2013; Colleoni & Arvidsson, 2014; Furnari, 2014; Gandini, 2015). On the other hand, such aspirations may often be at odds with the very nature of organizations. According to Santos and Eisenhardt (2005), boundaries reflect the very essence of the organization, the demarcation line between an organization and its environment, and the nature of boundaries can explain the success or failure of an organization. Boundaries necessarily relate to what is inside and outside the organization. The demarcation strategy chosen by each organization allow for efficiency, power, competence, and identity claims (Comeau-Vallée & Langley, 2019; Langley et al; 2019; Santos & Eisenhardt, 2005). To refer to this paradox, we refer to the concept of “boundary work.”

The general definition of “boundary work” refers to any effort to create, maintain, shape, or shift boundaries affecting groups, occupations, or organizations (Ashforth et al., 2000; Gieryn, 1983; Helfen, 2015; Lamont & Molnár, 2002). It is noteworthy that boundary work may be both about breaching and defending boundaries. The studies of Comeau-Vallée and Langley (2019), Quick and Feldman (2014), Meier (2015), Kellogg (2009), and Ungureanu and Bertolotti, (2018) highlight the possible coexistence of these opposite forms of boundary work within the same context: professional groups can create, maintain, or reconsider interprofessional and occupational distinctions in order to build new, more balanced social orders, oscillating between collaborative and defensive attitudes, and using available tools for such purposes.

The questions that rise are thus related to how organizations reconcile their need for internal-external demarcation and the attempts to create boundaryless spaces within their premises. Particularly, we are interested in understanding how a CS may help an organization manage internal and external boundaries, and whether they actually reach the purposes for which they were set up, considering their goals to become more innovative, on the one hand, and the need to maintain distinctiveness (i.e., efficiency, power, competence, and identity claims), on the other hand.

Context: Larnia Group’s design thinking CS

We conducted our case study in the headquarters of an industrial group that we fictitiously name Larnia, a leader in different segments of the food market. The Group excels in production and distribution strategies and invests in continuous product and process innovation. In the last years, Larnia has also started to embrace the paradigm of open innovation by building a CS in the headquarters. The purpose of the CS was to foster innovative cross-functional and multi-stakeholder projects guided by the design thinking approach. The CS was created following a “smart urban” style. It occupies a former factory building that was owned and operated by the Group. It is an open space, with minimalist design furnishings, often created with recycled materials. The space is available to all employees of the company, and also welcomes outsiders (e.g., customers, business partners, consumers, masters’ students, and researchers) who are invited to the design thinking sessions by the organizers of each session. Each group is called upon to solve a need or to optimize a product, or a process, within projects that last for different periods of time, from several weeks to more than a year. The space is managed by employees from Larnia’s R&D department.

Research design, data collection, and analysis: A grounded theory approach

We conducted a field study in Larnia and adopted a grounded theory research approach that entails iterations between data collection, data analysis, and theorizing (Strauss & Corbin, 1998). We conducted 30 semi-structured interviews. The semi-structured interviews were conducted face-to-face and have an average duration of one hour. The interviews were organized around a set of open-ended questions following an interview protocol on the management and use of the CS and the issues arising from its design and use. According to the grounded theory approach, the protocol was continually adjusted and changed during interviews, as to leave the interviewee free in providing us opinions and interpretations. Simple open question included: “Describe the CS”; “Why and when do you enjoy the CS”; “Describe a project in which you participated in the CS”; “Describe the use you have made or made of CS.”

Moreover, 23 interviewees were internal to the organization (coming from seven different functional areas) and seven interviewees were external to the organization, including actors from other organizations like external consultants and masters’ students. All interviews were fully recorded and transcribed. We obtained more than 1,000 pages of field notes that were used to elaborate the grounded model.

The data were analyzed following the grounded theory methodology and three phases of recurrent coding: open, axial, and selective (Strauss & Corbin, 1998). We used the Nvivo software to perform all the stages of coding. During open coding, we focused on the CS itself (features and design) and on the interpretations and uses that informants made of the CS. As we progressed through coding, it became obvious that boundaries were a recurrent theme in the data so we started refining our coding scheme to capture as many distinctions as possible regarding the ways in which informants talked about space and boundaries (e.g., how work boundaries were moved, overcome, or created in the exchange between users of the CS). It also emerged that informants often made distinctions between internal and external use of the CS, so we started thinking about internal and external boundary work, after going from data to theory and back, we came up with the three aggregate categories (described later) further refined our grounded model.

CS goals

To meet the innovation needs highlighted by the corporate strategy, Larnia decided to create a space in line with the open innovation paradigm, encouraging the search for innovation beyond organizational boundaries, and at the same time by promoting inter-functional projects with a high degree of diversity which aimed at fully exploiting the know-how and internal company skills. The CS, therefore, sets itself clear and defined objectives: to foster the generation of innovative ideas through the free exchange of knowledge between people with highly diverse backgrounds and to promote creativity inside and outside the organization.

The informants explained to us that Larnia designed the CS as a laboratory for innovation guided by the design thinking approach, a human-centered approach to innovation that draws from the designer’s toolkit to iteratively connect the needs of different stakeholders such as customers, consumers, business partners, with the possibilities of technology, and the requirements for business success (Brown, 2008). According to the design thinking paradigm, the sessions are organized in such ways that individuals with different backgrounds and/or cultures take part and feel engaged and fully motivated to bring a personal contribution, taking advantage of the enthusiasm that co-generation of ideas and contamination brings along (Brown, 2008).

During the data analysis, we found that all informants ideally referred to the CS as a laboratory in which it was possible to experiment new forms of aggregation and collaboration on inter-functional projects. According to the informants, the space as a laboratory had to be adaptive and located in a central location of the company for easy access. Additionally, it needed to be informal and equipped with tools to generate easy-to-use artifacts and prototypes that could further encourage the co-generation of ideas between people with very different skills and backgrounds. Physical features such as open spaces, bright colors, modular design, and informal workstations were expected to stimulate and favor free knowledge exchanges inside and outside the organization by tearing down organizational functions and hierarchies. Also, by adopting the design thinking approach, Larnia aimed at promoting creativity inside and outside the organization by encouraging co-creation and joint decision-making inside the CS.

The goal of CS on internal-external boundary work was therefore to help users break down barriers (in terms of goals, hierarchy, or skills) and modify their work boundaries, in order to use a new approach to innovation based on co-creation:

Implementation of CS goals

We found that despite the goal of using the CS as a laboratory for simultaneous internal and external innovation, Larnia’s CS generally hosted two different types of innovation sessions with different approaches: sessions for incremental innovations that featured internal collaborations between organization departments and sessions for radical innovations that featured collaborations between internal organizational departments and external stakeholders.

The design thinking sessions for incremental innovation with internal collaborations between organization departments were often organized by the members of the R&D department who invited the members of other organizational departments such as Marketing, Quality, Sales, and HR to participate in design thinking sessions organized for a well-defined project. The purpose of the design thinking sessions was often to generate innovative ideas about how to improve existing products or solve pre-existing problems to make product development and product-to-market processes more efficient. All sessions organized aimed at generating exchange and contamination of contents coming from different organizational areas, and at sharing different approaches and best practices for carrying out work activities.

Sessions for radical innovation are organized as collaborations between internal organizational departments and external stakeholders. These were typically composed of employees from internal departments – e.g., R&D, Marketing – who often launched the challenge and invited to the team external stakeholders such as retail clients, consumers, students, suppliers, and other stakeholders from the public arena as well, for instance, bloggers and journalists. The goal was to develop new idea for new products development. Our informants from Larnia often referred to this type of collaboration as “embedded,” as shown in the following excerpt:

Experienced boundary work during implementation

Experiences about boundary work in the CS refer to the way people actually used the CS to shape and shift their work boundaries, both internally and externally. In particular, we found that the CS afforded similar positive experiences (i.e., affordances) when used internally and externally, and different negative experiences (i.e., constraints) for internal and external innovation projects. Importantly, we show that the different goals that Larnia mobilized with respect to internal and external innovation (incremental and radical innovation approach, respectively), even if implemented in the same space, required different work method. Therefore, they mobilized users’ attention and resources differently such that path dependence was created – i.e., innovation problems were constrained by users’ subjective interpretations into pre-existing schemes available to the organization, shaping how internal and external collaborations were implemented, and generating a trade-off between the two.

Emergent constraints in internal boundary work

In addition to affordances for both internal and external boundary work, our evidence also highlights emergent constraints. Because each type of boundary work (i.e., incremental and radical innovation) was more effortful than initially expected, and required more resources, they became path dependent, i.e., individuals felt constrained by previous goals and experiences to treat the two separately. Consequently, informants experienced a trade-off between the ability to focus on internal boundary work and the ability to perform external boundary work.

Emergent constraints in external boundary work

Concerning external organizational boundary work, informants complained that external boundary work turned out to be much more complex than expected and, thus, also much more effortful.

Path dependence and CS: Trade-offs between external and internal boundary work

Managers of the CS repeatedly argued that the CS was a critical space which was difficult to animate with shared projects and populate with people with different background. The fact that the space was experimented as having different functions from the ones initially envisaged triggered users’ and space managers’ increasing efforts to make the CS more congruent to the initial goals of Cs as laboratory for innovation. Interestingly, although they dreamed of a CS that blurred internal and external boundaries and encouraged free co-creation, users continued to distinguish between the need to engage in incremental innovation with other internal functions and the need to open up to outsiders to perform radical innovation (i.e., strengthening path dependence). Importantly, as they struggled with the affordances and constraints generated by each type of innovation, users also explained that they needed more resources in terms of time and attention to perform each type of innovation boundary work successfully. Consequently, internal and external boundary works were perceived by our informants as competing and encroaching upon each other:

Based on this evidence, we concluded that in Larnia, there was a perception of path-dependent boundary work for innovation. Believing that internal boundary work was most suitable for incremental innovation projects and that external boundary work was most suitable for radical innovation projects, directing resources in one direction could come at the expense of the other. Importantly, the perception that internal and external innovations were path dependent was at odds with the initial goal of creating a CS that allowed to do both indiscriminately. This translated in an opposition between the goal of implementing the CS as a laboratory for free internal and external exchanges and the actual experiences that users had of the space – i.e., specialized, controlled, and chaotic area for incremental innovation with internal stakeholders and a black-boxed, restricted showcase for radical innovation with external stakeholders. In other words, what users had meant as a simplification triggered additional complications in Larnia’s management of innovation projects. Specifically, our informants reported that the multiplication of the CS’s affordances and constraints with respect to Larnia’s initial goals led to the perception that the space was becoming increasingly complex and heterogeneous, and that interacting in the CS required increasing relational efforts which could be difficultly sustained internally and externally at the same time.