In today's business environment, firms keep searching for new and effective innovation models, of which the so-called ‘open innovation’ approach seems to be largely spread (Alberti, Giusti, Papa, & Pizzurno, 2014; Alberti & Pizzurno, 2013).

While a large part of the available literature has concentrated the attention on large innovators (Van de Vrande, de Jong, Vanhaverbeke, & de Rochemont, 2009), academic interest in this topic has now shifted on other kinds of innovators and makers actually do qualify for playing a crucial role in open innovation networks.

In Makers: The New Industrial Revolution, Chris Chris Anderson (2012), the former editor-in-chief of Wired, describes how open-source design and 3D printing “bring manufacturing to the desktop”. Much of this activity has been fuelled by a new generation of hardcore makers, who have come together under the banner of the “Maker Movement”. The Maker Movement, whose journey “from the margins to the mainstream” was chronicled in the March 2011 Wired cover story, is deeply committed to local, collaborative, community-based invention; makers network to share tools, technologies, and ideas through online communities but also through physical spaces. The MIT pioneered this field setting up a Center for Bits and Atoms (http://cba.mit.edu/) and a fab lab (http://fab.cba.mit.edu/) that aim at an interdisciplinary spreading of material open production in a wide range of academic and economic fields (Thompson, 2005).

As Maker communities spring up around the globe, a variety of physical and virtual platforms to serve them have emerged – from platforms that inspire and teach, to those that provide access to tools and mentorship, to those that connect individuals with financing and customers (Deloitte, 2014; Gershenfeld & Euchner, 2015). Several concepts compete for understanding the makers communities: knowledge networks, knowledge communities, communities of interest, communities of practice, etc. (e.g. Brown & Duguid, 1991; Lave & Wenger, 1991). These resemble both virtual platforms like Instructables, Makerspace.com, etc. as well as locally based communities of makers either temporary (e.g. Bay Area Maker Fairs), in the form of temporary clusters (Belussi, Sedita, & Omizzolo, 2008; Maskell, Bathelt, & Malmberg, 2004), or stable (e.g. Central Florida makers community), with related and supporting industries and institutions in the form of clusters (Lazzeretti, Boix, & Capone, 2010; Porter, 2008), where knowledge is co-constructed and shared, through joint problem-solving. A preliminary review of the literature, drawing on previous studies on the clustering of creative industries – like the Makers Movement – suggests that several factors are at play to explain why collaborative knowledge networks or communities of practice might cluster (e.g. Lazzeretti et al., 2010).

On a physical level, the rise of the maker culture is closely associated with the rise of a totally new entrepreneurial ecosystem (Mortara & Parisot, 2016) made of hackerspaces, fab labs, makerspaces, tech-shops, co-working spaces, crowdfunding platforms, related and supporting industries (first of all laser-cut and 3D printers makers and consultants, but not only; dedicated vocational training and education, academic and corporate research, etc.), local and international associations, clubs and institutions.

On a virtual level, sharing and cooperation in the Maker Movement is also facilitated by Internet-based platforms (Chesbrough, 2003). Internet-based sharing, marketing, distribution, and crowd funding (such as platforms like Kickstarter, Indiegogo, and FundedByMe) have given makers access to resources hitherto unavailable through traditional physical platforms (Mollick, 2014).

The makers’ culture is experimental and open (Shah, 2005). Collaboration and sharing are so essential to the Makers Movement that a survey titled Maker Market Study (Make, 2013) which examined the extent of makers’ sharing and collaboration reported that 59% of all respondents either said “other use what they make” (41%) or they “make things with others” (50%). Collaborations and interactions are the backbone of the maker culture (Von Busch, 2013). The maker culture is thus less of a do-it-yourself (DIY) and more a do-it-together culture, merging collaborative play and interactions, often for the sake of shared curiosity.

Recent studies (see Alberti & Pizzurno, 2015 for a comprehensive review) on open innovation have shown that successful innovation does not depend exclusively on technological knowledge but rather on a heterogeneous recombination of a broader set of knowledge types (Boschma & Ter Wal, 2007; Rodan & Galunic, 2004). Knowledge is a broad concept, encompassing various types (Amin & Cohendet, 2004; Simonin, 1999) such as technological knowledge (i.e. product and process development knowledge), market knowledge (i.e. organized and structured information about the markets), and managerial knowledge (i.e. competencies to make managerial and organizational processes more efficient and effective), as well as industry-specific knowledge (i.e. idiosyncratic competencies that strictly refer to a specific sector), financial knowledge (i.e. access to finance and competencies on how to deal with financial players) and institutional knowledge (i.e. translation of historical data, traditions, values and norms).

Knowledge exchange can offer a series of benefits, such as reducing time to market, cost and risk reduction, improving access to specific expertise, etc.

The combination of the aforementioned streams of literature in open innovation – (i) open innovation and makers and (ii) open innovation and knowledge exchanges – suggest that the research gap in literature is twofold.

When studying knowledge flows and exchanges in innovation networks, some authors (Alberti & Pizzurno, 2015) have suggested to dig more in their multiplexity (Gimeno & Woo, 1996), i.e. the concurrent and not-intended exchange of different forms of knowledge (i.e. market, managerial and technological knowledge) at the same time. The idea is that different types of knowledge may be embedded one in the other or that the exchange of one type of knowledge may deliberately or unconsciously imply the exchange of other types of knowledge. The multiplexity in knowledge exchange is dramatically crucial in open innovation networks, where, despite the setting is open, the intent is to govern which actor is bringing in/out a specific knowledge contribution for the sake of innovation.

Hence, our research question is the following.

• RQ: Whether and how knowledge leaks occur in open innovation networks with makers?

To answer our research question, the research design of this study relies on Social Network Analysis (SNA) methods and techniques. SNA has been increasingly used in management science and business studies (Borgatti, Mehra, Brass, & Labianca, 2009) and – more specifically – to investigate open innovation networks and knowledge exchanges (Alberti & Pizzurno, 2015; Cantner, Meder, & Ter Wal, 2010; Giuliani & Bell, 2005; Pe’er & Keil, 2013; Sammarra & Biggiero, 2008).

SNA data collection and analysis allowed us to elaborate a preliminary discussion on the practices of open innovation and the specific role of makers. First, we disentangled the role of makers in open innovation networks vis-à-vis the three types of knowledge considered in this study, i.e. market, managerial and technological knowledge. Second, we tested whether knowledge flows were affected by multiplexity – i.e. the concurrent flow of multiple types of knowledge, even unconsciously. Finally, SNA techniques allowed us to confirm the occurrence of ‘knowledge leaks’ in open innovation networks with makers.

We followed previous studies applying SNA to the study of innovation networks (e.g. Alberti & Pizzurno, 2017) and we employed a snowball sampling technique to collect data (Frank, 2005). We started our sampling from a list of makers made available by a makerspace operating in the selected cluster, namely Monzamakers. Players in the initial sample were then asked to nominate other players involved in their practices of open innovation with firms. And we re-iterated the process until when makers, fab labs and firms named as being part of open innovation practices were always the same. The final sample is the one reported in Table 1.

Interviews were based on a structured questionnaire and data collected have been turned to anonymous to protect sensitive data, considering unintended drawbacks and implications of disclosing innovation practices. Data collection on the cluster and the community of makers in the cluster relied on secondary data made available both from the cluster organization and the main fab labs/makerspaces in the area.

Data collection allowed the construction of four different networks. First, interviewees were asked to list all other actors with whom they engaged in open innovation practices in the previous year. Each one of the actors mentioned during interviews was then complemented with data on their location and their profile. Finally, we asked to evaluate each collaboration in terms of market, technological and managerial knowledge received and transferred (see Table 2).

Giuliani and Bell (2005), who use the cognitive position of players combined with their degree of openness, distinguished among gatekeepers (i.e. players with a central position in the network in terms of knowledge transfer to other proximate players and strongly connected with external sources); mutual exchangers (i.e. players that have a balanced source/absorber position); external stars (i.e. players that have established strong linkages with external sources, but have limited links with local knowledge system); isolated actors (i.e. players that are poorly linked in general). Accordingly, we measured the cognitive position of each player as well as its degree of openness, to define four types of actors. The cognitive position was obtained by calculating the ratio between the in-degree and out-degree for each actor in all the three knowledge networks considered. The degree of openness was instead measured calculating the proportion of the knowledge exchanged at the local level with respect to the total knowledge one particular actor exchanges. This value was then subtracted from 1. We, then, defined four types of roles played by actors in the knowledge system: (a) gatekeepers (cognitive position with an in/out degree centrality <1 and an external openness above the average); (b) mutual exchangers (cognitive position with in/out degree centrality about 1 and their external openness is on average); (c) external stars (cognitive position with an in/out degree centrality >1, but also a very high external openness); (d) isolated firms (cognitive position with an in degree centrality as well as an out degree centrality of about 0 and external openness above the average).

Finally, we discriminated by four types of cluster actors in our study: firms (either small-sized or large corporations), makers, research organizations (research centres, laboratories and universities) and fab labs or makerspaces.

For network data analysis, we relied on SNA techniques and we used UCINET 6 (Borgatti, Everett, & Freeman, 2002) as a software package.

First, we constructed the entire open innovation network (binary, symmetric and not oriented) and the three knowledge networks – technological, managerial and market knowledge (valued, symmetric and oriented) and we described them in terms of isolates, average degree (out/in) and components. Then, we focused solely on players involved in either technological or market or managerial knowledge exchanges in open innovation networks.

To answer our research question, we then applied Giuliani and Bell (2005) typology to dig into the role of makers in open innovation networks and then – following Fonti (2002) – we checked for multiplexity in knowledge flows, to explore knowledge leaks in open innovation networks with makers.

First, we analysed the overall network emerging from the answers obtained from the interviewed organizations. The open innovation network obtained is reported in Fig. 1, while the exchange of technological, market and managerial knowledge in the cluster in Figs. 2–4 respectively. Actors are distinguished by type: smallest black nodes represent makers, other black nodes (depending on their size) represent small-sized firms (medium node) or large corporations (large node) respectively; research organizations are in white and fab labs and makerspaces in grey.

Fig. 1.

The entire open innovation network.

(0.04MB).

Fig. 2.

Technological knowledge exchanges.

(0.04MB).

Fig. 3.

Market knowledge exchanges in the cluster.

(0.03MB).

Fig. 4.

Managerial knowledge exchanges in the cluster.

(0.03MB).

Consistently with previous literature (Giuliani, 2007a, 2007b; Sammarra & Biggiero, 2008) our preliminary investigation of open innovation networks confirms that the three types of knowledge considered are unevenly exchanged. The exchange of technological knowledge seems to be open to most of the actors, whilst the exchange of market and managerial knowledge appears as more selective. As reported in the paper by Alberti and Pizzurno (2015), different types of actors have also different roles in open innovation networks, especially with reference to the three types of knowledge considered in this study.

Table 3 confirms what emerges from visual representations. The technological knowledge network involves the vast majority of the actors and firms exchange more technological knowledge than other types. The average degree of the technological knowledge network is indeed significantly higher than in the other two cases. This aspect is also reflected in the number of components, that shows how market and managerial knowledge networks are much more fragmented.

With the specific aim to disentangle the role of makers in open innovation networks vis-à-vis the three types of knowledge considered, we deepened our understanding of the different knowledge brokerage roles played by makers and firms (either large corporations or small-sized firms) applying the typology suggested above.

Fig. 5 positions actors in the managerial knowledge network. Using Giuliani and Bell's taxonomy (2005) there are several gatekeepers (low in/out-degree ratio and high openness) who are mainly large corporations and some small-sized firm with connections outside the local context, who absorb inputs from the global industry and then transfer them locally. There are also some mutual exchangers and a few external stars which are actors with several international connections. They are often makers that are very specialized on specific technological applications or fields. Finally, there are just a few isolated firms.

Fig. 5.

Firms’ roles in managerial knowledge exchanges.

(0.06MB).

With regard to market knowledge networks (Fig. 6), there is a relative presence of gatekeepers. It seems therefore that only a limited number of actors have access to global market knowledge networks, while the others are absolutely excluded from a direct access to it. Interestingly, the vast majority of these organizations are makers that tap into global networks of peers with the same interests, expertise and focus. There is also a significant number of mutual exchangers, i.e. firms that have a balanced source/absorber position.

Fig. 6.

Firms’ roles in market knowledge exchanges.

(0.07MB).

For technological knowledge exchanges, there is a more heterogeneous and not so clear situation. A variety of roles are present (Fig. 7), for example some makers seem to be acting as external stars while just a few perform the role of gatekeepers. The large corporations are positioned more in the role of absorber or mutual exchanger, consistently with their role in the supply chain where they act as integrators of sub-components and organizers of the supply chain.

Fig. 7.

Firms’ roles in technological knowledge exchanges.

(0.07MB).

In conclusion, our first analysis revealed that in open innovation networks, whilst in managerial knowledge exchanges large corporations and some small-sized firms act as gatekeepers and makers as external stars, in market knowledge makers are gatekeepers and bring in the knowledge they access from global niche markets. This also happens in technological knowledge exchanges where large corporations absorb and makers act as gatekeepers. Such a composite situation, suggested us that while when firms engage in open innovation practices they do absorb from makers market and technological knowledge, it is makers who benefit from larger firms when it comes to managerial knowledge. This implies – as already suggested in literature (Alberti & Pizzurno, 2015) – that knowledge flows are affected by multiplexity.

Hence, we checked for multiplexity in knowledge exchanges, exploring how the different knowledge networks are interrelated. At first, we performed a correlation analysis of the different knowledge networks. We employed a Quadratic Assignment Procedure (QAP), a test based on the repetition of a permutation for an arbitrarily large number of times (De Lange, Agneessens, & Waege, 2004; Kilduff & Krackhardt, 1994). Results of the correlation test are presented in Table 4. All the results, obtained running 10,000 permutations, are significant (p<0.001) and show a very strong correlation among the three types of knowledge exchanges.

Next, we conducted a more fine-grained analysis of the simultaneous exchange of different types of knowledge through a multiplexity analysis (Fonti, 2002). Table 5 presents some basic information about the knowledge networks resulting from the combination of the exchange of technological and/or market and/or managerial knowledge. The test shows how different combinations are less likely to occur than others and involve fewer players. Moreover, the exchange of managerial and market knowledge takes place less frequently than other exchanges and when it happens, it is always associated with the exchange of technological knowledge as well. In both cases, the number of isolates is higher than in other knowledge combinations. This can be the consequence of the fact that not all players are involved in the simultaneous exchange of multiple knowledge as well as to the specific nature of some organizations or to their limited absorptive capacity, which represents a threshold to the knowledge they can receive from partners. Hence, actors are often engaged in simultaneous exchanges of knowledge.

In conclusion, we verified whether some simultaneous knowledge exchanges are more likely to happen than others. We applied a bootstrap technique suggested by Snijders and Borgatti (1999) to compare the densities of the different networks and understand whether such differences are the result of pure random chance or not. The results of this test are presented in Table 6.

Interestingly, the difference in density takes a meaningful and significant value only when comparing one network including the exchange of technological knowledge with a network not including it. Again, the relevance of the exchange of technological knowledge emerges, suggesting that technological knowledge exchanges may imply even the unintended exchange of either market or managerial knowledge, configuring what in literature is referred to as a ‘knowledge leak’ (Alberti & Pizzurno, 2017). Furthermore, the exchange of the three types of knowledge at the same time is constantly less frequent than other multiple exchanges, confirming the selective nature of knowledge flows.

This study is positioned in one of the hotspots of research on innovation, where open innovation studies (West et al., 2014), account for knowledge exchanges (e.g. Alberti & Pizzurno, 2017) in the field of makers (e.g. Aldrich, 2014; Browder et al., 2017; Van Holm, 2015, 2017; Von Busch, 2013), through the lenses of SNA (Ter Wal & Boschma, 2009).

First of all, our findings confirm multiplexity and heteromorphism in knowledge exchanges (e.g. Biggiero & Sammarra, 2010; Boschma, 2005). Open innovation networks are the result of the exchange of multiple types of knowledge by a variety of actors that play heterogeneous roles vis-à-vis different types of knowledge.

Second, the present study sheds light on a totally unexplored phenomenon, that is the phenomenon of open innovation and makers, focusing on knowledge leaks. Makers are eager to absorb knowledge from other players, when they engage in open innovation practices. Makers engaging in open innovation networks may largely benefit from absorbing several types of knowledge either on technology dynamics and applications, market scenarios and evolutions or managerial practices and requirements. This is confirmed by our empirical evidence, according to which makers act as external stars, i.e. not sharing their specialized technological expertise but absorbing general technological knowledge from the cluster ecosystem (Alberti, Sciascia, Tripodi, & Visconti, 2011) and, given the evidence that technological knowledge is inextricably linked to market and managerial ones, by multiplexity effect, open innovation networks become a gateway to several knowledge leaks that makers absorb. Conversely, our work sheds light on the fact that makers convey knowledge flows to larger and consolidated firms. Our evidence confirms the general understanding that makers may be a rich source of knowledge leaks. Typically, makers operate in global niches, where dedicated technological knowledge is inextricably linked to a certain kind of market knowledge. Given their credo and their immaturity in managerial processes and in open innovation practices and their eagerness in collaborating with key players in the field, for innovation-specific matters, they might pay less attention in engaging in open innovation networks and succumb to knowledge leaks to the benefit of their partners who can access through them particular knowledge.

The exploration of open innovation practices with makers allowed us to understand, consistently with previous studies (Alberti & Pizzurno, 2017), how one type of knowledge – in our case, the technological one – may be conducive of other types of knowledge, qualifying as a platform or a language that drives market and managerial knowledge leaks.

Our results have practical implications both for policy makers and for managers. First of all, our research confirms how open innovation often originates from a combination of different knowledge types acquired through the collaboration with heterogeneous players, makers included. Hence, managers may design open innovation strategies balancing their portfolio of collaborations to maximize the absorption of relevant knowledge and makers may consider engaging in open innovation practices to accelerate knowledge absorption and eventually launch a start-up. Nevertheless, our study warns managers against the risk of knowledge leaks, especially in those cases where the eagerness or the prestige to take part to some open innovation networks may hamper them to protect or to control over knowledge leaks. This opens up for possible interventions for policy makers too. First of all, policy makers may consider incorporating the concept of knowledge leaks in their campaign in favour of open innovation. Second, our study may help policy makers in designing programmes for knowledge transfer partnerships between firms and makers in a more conscious way, especially warning new to business players, like makers, about possible leaks, despite their culture of openness. Finally, there is also the need of developing professional figures like consultants, maybe operating in fab labs, co-working spaces or makerspaces that can mentor and advise makers on how to get engaged in open innovation networks and, hence, find their specific role in the cluster.

The present paper has some limitations that open avenues for future research. Given the explorative nature of this paper that looks at a totally unfamiliar phenomenon, there is abundant room to go in-depth in the refinement of the concept of knowledge leak, even operationalizing it for future tests. Future studies on this line, may even consider developing a specific set of SNA measures for the specific analysis of knowledge leaks to complement the ones on multiplexity. Future studies may continue to explore the phenomenon through in-depth case studies on both sides.

Despite these limitations, our data illuminate several aspects of whether and how knowledge leaks occur in open innovation networks with makers, opening up intriguing aspects that have been overlooked by extant literature and triggering several lines of further research on the topic.