Abstract
This paper aims to provide a new perspective on the collaboration between Micro, Small, and Medium Enterprises (M&SMEs) and universities. It seeks to address a gap in the existing literature, which predominantly focuses on the analysis of collaboration between large firms and universities. This partnership is widely regarded as mutually beneficial, as M&SMEs can gain a competitive edge through innovation, while universities could leverage research outcomes and refine their academic programs. The study adopts a systematic literature review methodology and conducts a thematic analysis of articles obtained from the Scopus and Web of Science databases, employing an integrated interpretative framework based on six different perspectives: (1) who (involved actors); (2) what (collaboration inputs and outputs); (3) when (analyzed time span); (4) where (countries); (5) how (implemented collaboration modes); and (6) why (drivers, enabling factors and barriers). A total number of 82 journal articles published between 1987 and 2023 were analyzed. Based on the adopted interpretative framework, a lack of a shared conceptualization of the investigated phenomenon emerged and is associated with a relatively scarce adoption of theoretical frameworks. Moreover, a highly differentiated set of barriers and enabling factors were found, even if their interdependencies appear rarely investigated.
Keywords
Introduction
Powell and Snellman (2004) were among the social scientists highlighting that, in the last decades of the past century, the most advanced industrial economies shifted from being based on natural resources and physical inputs to being grounded in intellectual assets. Based on this evidence, they developed the concept of the “knowledge economy”, defining it as “production and services based on knowledge-intensive activities that contribute to an accelerated pace of technical and scientific advance” (p. 199). Within this context, the collaboration between universities and companies—generally referred to as University-Industry collaboration (U-IC)—assumes a special role (El-Ferik and Al-Naser, 2021), resulting in a win-win relationship (Pujotomo et al., 2023). More specifically, this collaboration enables both actors to reach their strategic aims and pursue their missions. For instance, firms can build on breakthrough technologies often developed by universities, compensating for their lack of in-house knowledge. At the same time, universities may exploit the results of their research activities and fine-tune their academic curricula to better match firms’ demands for technical, technological, and managerial competences (Al-Tabbaa and Ankrah, 2016; Lee and Win, 2004).
Moreover, U-IC has beneficial effects on the social and economic system, as it may boost economic growth (Pujotomo et al., 2023) and sustainable development (Bengoa et al., 2021; Murdiati et al., 2023). Despite the increasing advantages generated by U-IC, this partnership has not yet achieved its full potential (Figueiredo and Ferreira, 2022).
As Alves et al. (2015) suggest, government incentives are crucial for overcoming the obstacles that continue to impede this relationship. Consequently, governments are encouraging higher education institutions to diversify their revenue streams by commercializing research through technology licensing and the creation of spin-off companies (Mosey et al., 2007).
Although many are skeptical about the role of governments (Figueiredo and Ferreira, 2022), they are crucial in U-IC, particularly through public funding that supports research and private sector development (Badillo et al., 2017). The Triple Helix model of collaboration (Etzkowitz and Leydesdorff, 2000) is essential for firms’ economic growth and strategic advantage. According to this model, each actor evolves from its initial role (namely, basic research for universities, production and commercialization of goods for companies, and regulation for governments) toward a certain form of hybridization. This, in turn, allows the creation of new intermediaries among the three actors, such as University Technology Transfer Offices (TTOs), which transform university knowledge with commercial value into marketable goods. The hybridization of university and industry—traditionally conceptualized as two distinct business models (Buehling and Geissler, 2022)—has boosted the rise of entrepreneurial universities and even open innovation (De Bernardi et al., 2020).
Building on this evolution, it becomes evident that although U–IC can in principle be implemented by any type of company, the majority of existing studies have focused on large firms (Parmentola et al., 2021; Pereira and Franco, 2023). However, such companies are no longer the only ones collaborating with universities (Rantala and Ukko, 2018). Generally, Micro, Small, and Medium Enterprises (M&SMEs) (defined as companies having, respectively, fewer than 9, 50, and 250 employees; European Commission, 2020) often prefer to develop innovations through interactions with suppliers and providers (Garcia-Perez-de-Lema et al., 2017). An increasing number of these firms are activating links with universities to create an adequate competitive advantage (Bishop et al., 2011; Perkmann and Walsh, 2007).
Recently, Teirlinck et al. (2022) pointed out that M&SMEs may prefer collaborating with universities rather than customers and suppliers, as universities are less interested in exploiting the partner’s knowledge. Moreover, such a collaboration may allow M&SMEs to develop either a short-term advantage, as in the case of a single innovation (Messeni Petruzzelli and Murgia, 2023), or a long-term sustainable competitive advantage (Bishop et al., 2011). At the same time, M&SMEs differ significantly from larger companies in terms of innovation processes (Çakar and Ertürk, 2010), as they are generally characterized by liabilities of smallness, newness, and connectedness (Pereira and Franco, 2022). They also face limitations in financial and human resources (Lin and Yang, 2020), which are essential for their absorptive capacity (Cohen and Levinthal, 1990; Dornbusch and Neuhäusler, 2015; Teirlinck et al., 2022). This, in turn, is considered a necessary condition to transform the results of a collaboration into an innovation (Apa et al., 2021). Although M&SMEs are agile and responsive, their weak managerial and technical skills, along with limited financial resources, can impede effective technology transfer and may deter higher education institutions from engaging in technology transfer projects (Tang et al., 1996). Furthermore, Fransman (2008) states that the costs and benefits of U-IC are influenced by the size of the firms involved.
Finally, M&SMEs are quite a heterogeneous category; therefore, a one-size-fits-all solution is generally not effective (Ranga et al., 2008). Thus, the collaboration between M&SMEs and universities emerges as a topic worth investigating as a specific target within the wider field of study concerning U-IC.
In this respect, Pereira and Franco (2022) proposed a structured literature review aimed at investigating the characteristics of the collaboration between Small and Medium Enterprises (SMEs) and universities. However, their investigation only considers journal articles published between 1995 and 2019. As recently pointed out by Bengoa et al. (2021), topics such as U-IC and technology transfer are increasingly attracting the attention of scholars; therefore, “periodic literature reviews are necessary to recompile and synthesize the topics studied” (p. 1542). Moreover, Snyder (2019) pointed out that when new publications are added to the existing literature on a specific research field, it becomes more fragmented and interdisciplinary, requiring an updated assessment of the state-of-the-art. Finally, the analysis conducted by Pereira and Franco (2022) was mainly focused on bibliometric issues (e.g., most cited articles, most productive scholars) and offered a general description of the collaboration characteristics but did not specifically define variables affecting the process (e.g., drivers, enabling factors, barriers). Therefore, this paper aims to provide an updated qualitative and quantitative analysis of the existing literature on the relationship between universities and M&SMEs. To offer a comprehensive understanding of the investigated phenomenon, a holistic approach is assumed based on the so-called “six Ws” methodology (Who, What, When, Where, hoW, and Why) suggested by Callahan (2014). This methodology, often referred to as “5W and 1H,” has been adopted to develop structured literature reviews regarding management topics (see, for instance, Di Stefano et al., 2023; Xie et al., 2017; Garg and Sushil, 2023). More specifically, the following six Ws issues are considered and captured in the analysis. (1) (2) (3) (4) (5) (6)
In order to analyze these issues, the rest of the paper is divided into four main sections, the first of which offers a detailed description of the methodology adopted to implement the structured literature review. The following section contains findings articulated according to the six Ws described earlier. After this, methodological recommendations and avenues for future research are proposed. Concluding remarks, implications –for university managers, M&SMEs entrepreneurs and policymakers, respectively– and limitations are summarized in the last section.
Methodology
The paper aims to shed new light on the collaboration between M&SMEs and universities according to the six issues presented earlier. More specifically, the authors develop a systematic review of the existing literature, defined as “a systematic, explicit, and reproducible design for identifying, evaluating, and interpreting the existing body of recorded documents” (Fink, 2005: p. 6). To achieve this aim, the authors adopted the Seuring and Gold (2012) process model for content analysis, which is divided into four main steps. The first step, referred to as “material collection,” involves gathering the documents to be analyzed in the following steps. In this regard, the most recent structured literature reviews on U-ICs (independent of firm size; Pereira and Franco, 2022; Pujotomo et al., 2023) and the broader topic of technology transfer (Bengoa et al., 2021) used a unique dataset, selecting data from either Elsevier Scopus or Web of Science (WoS). However, following Martín-Martín et al. (2018), we decided to consider both datasets to avoid omitting documents with high impact. We decided not to use Google Scholar since it contains almost all citations found by WoS (95%) and Scopus (92%); moreover, the remaining large number of unique citations are not from journals and, on average, have a much lower scientific impact than citations found by the other two datasets (Martín-Martín et al., 2018). We considered journal articles published until December 2023 and written in English (referring to the entire article, not only the abstract). The search strings used to retrieve documents are listed in Appendix 1.
Based on this approach, 278 documents were found (172 from the Elsevier Scopus dataset and 106 from WoS); however, 71 were duplicates (19 from Scopus and 52 from WoS) and were therefore eliminated. After this, the following exclusion criteria were adopted. (a) papers containing only a review of the extant literature (e.g., McCullough, 2003; Pereira and Franco, 2022); (b) articles regarding research centers other than universities (e.g., Puliga et al., 2020), since the latter differ from other public research institutions (Liu et al., 2020). Due to these differences, articles analyzing both institutions without splitting findings among them were also eliminated (e.g., Pinto et al., 2015); (c) articles referring to both M&SMEs and large companies when findings were not split among the different categories (e.g., Pinto et al., 2015; Serbanica et al., 2015); (d) documents regarding collaborations where the M&SMEs do not directly cooperate with a university, as a third actor (e.g., scientific park) intermediates the relationship (e.g., Nishimura and Okamuro, 2011); (e) documents regarding technology transfer activities consisting merely of the establishment of university spin-offs. However, collaborations between universities and already established university spin-offs were included.
Sampled documents according to the adopted selection process.
Source: own elaboration.
The second step of the Seuring and Gold (2012) process model concerns the “descriptive analysis”. Therefore, the authors analyzed the formal characteristics of the selected documents. In this regard, the data summarized in Figure 1 show that scholars’ interest has significantly increased since 2013, with the partial exception of 2019, thereby confirming the importance of an updated structured literature review. Breakdown of sampled documents by year. Source: own elaboration.
Breakdown of sampled documents by journal.
Source: own elaboration.
Breakdown of sampled documents by theoretical perspectives.
Source: own elaboration.
Breakdown of sampled document by adopted research typology and methods.
Source: own elaboration.
A final characterization of the sampled articles is regarding the presence of definitions of the collaboration between universities and M&SMEs. The authors found that only seven out of the 82 chosen documents offer a clear definition of the investigated phenomenon. This finding is quite similar to the one reached by Bengoa et al. (2021), which investigated 3218 journal articles analyzing the wider phenomenon of technology transfer, and includes the one investigated in this article. They explained such a limited result, stating the concept (technology transfer) is “a complex, difficult process that also needs time to evolve” (Bengoa et al., 2021). Such an interpretation could also be applied to the collaboration between M&SMEs and universities, given the variety of modes that can be implemented, as will be demonstrated when discussing the findings related to the ‘How’ dimension.
The third step of the analysis was “category selection”, which consists of defining the analytical categories to classify documents’ contents. As already pointed out in the Introduction, the authors adopted the “six questions approach” proposed by Callahan (2014), namely “Who”, “What”, “When”, “Where”, “Why” and “How”. In Figure 2, the adopted interpretative framework is synthesized. In order to operationally describe the contents of each issue, a short description of the investigated variables is provided as follows. (a) Who: refers to the actors involved in the collaboration (universities vs M&SMEs); (b) What: pertains primarily to the nature of inputs (e.g., the nature of the transferred/developed knowledge/technology) and outputs (e.g., typology of the innovation deriving from the transfer) characterizing the collaboration. Based on such analysis, the directionality of the relationships is classified, making it possible to find two main alternatives, namely unidirectional – as in the case of contract research commissioned by an M&SME to the university – and collaborative/joint agreement (Capaldo et al., 2016) – when the research project requires the engagement of both partners; (c) When: concerns the time span analyzed by authors, enabling verification of the existence of longitudinal studies as well; (d) Where: focuses on the geographical dimension of the sampled studies, facilitating examination of the presence of comparative studies; (e) How: is related to the modes adopted by universities and M&SMEs to implement the collaboration and their possible classifications; (f) Why: looks at three different issues. The first is related to factors inducing universities and M&SMEs to collaborate (drivers), the second refers to elements facilitating the relationships (enabling factors) and the third one concerns those hindering it (barriers). The adopted interpretative framework. Source: own elaboration.
The final step of Seuring and Gold’s (2012) process model for content analysis is regarding material evaluation. This activity was performed by reading, analyzing and coding all selected documents according to the six issues presented earlier. The process reliability was improved by discussion among authors (researcher triangulation) and by ensuring process documentation (Denyer and Tranfield, 2009).
Findings
The Who issue
Typologies of investigated universities and actors within them.
Source: own elaboration.
Typologies of investigated M&SMEs.
Source: own elaboration.
Breakdown of investigated M&SMEs by sector and industry.
Source: own elaboration.
The what issue
Directionality of collaboration.
Source: own elaboration.
Breakdown by collaboration input.
Source: own elaboration.
Breakdown by collaboration output.
Source: own elaboration.
The when issue
Breakdown by time span extension.
Source: own elaboration.
The where issue
Breakdown of empirical sampled documents by country.
In addition to the national-level distribution, a significant stream of research highlights the relevance of regional and spatial dimensions in shaping U-ICs. Foundational studies on regional innovation systems (Asheim and Gertler, 2005; Cooke, 1992) and clusters (Porter, 1998) demonstrate that innovation processes are often territorially embedded, relying on geographical proximity, local networks, and institutional thickness. From this perspective, collaborations emerge not merely as the result of national policies or aggregate indicators, but rather as outcomes of localized interactions between universities, firms, and other stakeholders. More recent contributions on regional innovation ecosystems (Autio et al., 2014; Guerrero and Urbano, 2019) and place-based policies (Foray et al., 2011; McCann and Ortega-Argilés, 2015) further reinforce the idea that geographical context plays a decisive role in shaping both the opportunities and constraints of U-ICs. Consistent with this line of inquiry, our findings indicate that 23 out of the 82 papers specifically address regional areas within the analyzed countries. Notably, such studies began to appear only in the late 1990s and became more widespread over the last two decades. At the same time, research focusing on rural contexts remains extremely limited, with only one paper explicitly examining U-ICs in these areas (Johnston and Prokop, 2021).
The How issue
The How issue refers to the modes of interaction that universities and M&SMEs use during their collaboration. Scholars identified 23 different alternatives, ranging from MSc and/or PhD theses to co-design projects. Therefore, it seems useful to classify them into homogeneous categories. In this regard, very few authors proposed specific classifications; moreover, such proposals differ in terms of adopted criteria. However, three main typologies of these criteria seem to emerge. (a) Collaboration aims: this classification criterion is regarding aims inducing M&SMEs and universities to collaborate. Such a criterion was initially proposed by Perkmann and Walsh (2007) who identified the following two alternatives: (i) “academic engagement” – where the receiver aims to have access to the university base of knowledge (e.g., contract and/or collaborative research, training, consultancy); and (ii) “commercialization” – where the receiver aims to have access to university research outputs (e.g., through patents sale and/or licensing). In other words, while “academic engagement” is a form of knowledge-related collaboration between partners (Bozeman et al., 2013), “commercialization” represents a sort of “market acceptance” for academic research results (Capaldo et al., 2016; Markman et al., 2008). The classification under investigation may be considered similar to the one proposed by Corral De Zubielqui et al. (2015), which differentiates between university-industry links and relationships. While in the former the receiver aims to access university resources and research outputs and/or activate human mobility transfer (e.g., graduate recruiting), the latter is regarding the development of new research outputs. Therefore, while “relationships” are consistent with the “academic engagement” alternative, “links” refer to the “commercialization” one. In contrast to this perspective, Marullo et al. (2022) affirm that the two forms of U-IC could be strictly interconnected, with positive interaction effects between academic engagement and commercialization in fostering research commercialization outcomes. Finally, Jones and Corral De Zubielqui (2017) propose differentiating between “generic links” – which is regarding knowledge/technology transfer – and “relational links” – which includes either research services or research partnerships. Therefore, while generic links are more similar to the commercialization option, the relational links fit better with the academic engagement alternative; (b) Content-based: such a criterion concerns the content of the relationships between universities and M&SMEs. In this regard, Brimble and Doner (2007) identified three main categories, namely: training and education, consultancy services, and research. In contrast, Pinto et al. (2015) offer a more articulated set of contents, including: (i) advanced services (e.g., consultancy and contract research); (ii) collaborative research (e.g., joint R&D projects); (iii) human resource-based activities (e.g., internships for graduates, exchange of personnel, training for employees); (iv) commercialization activities (e.g., patent exploitation, participation in spin-offs); and (v) informal relationships; (c) Degree of formalization/Governance mode: Garcia-Perez-de-Lema et al. (2017) differentiated the collaborations between universities and M&SMEs according to the mode by which they are governed. More specifically, they propose two alternative approaches, namely: contractual-based and relational-based. While the former typology includes collaboration modes such as research conventions, R&D development projects and innovation ones, the latter includes business alternatives such as business and/or technical training, consultancy and fellowships for students. More recently, Apa et al. (2021) suggested differentiating between collaborations according to the degree of formalization, i.e., differentiating between formal and informal ones. The formal category includes “personal formal collaborations” (e.g., student internships and use of university facilities), “formal non-targeted agreements” (e.g., endowed chairs and advisory boards), and “formal targeted agreements” (e.g., patenting and licensing agreements and cooperative research projects). In contrast, informal collaborations include a set of personal informal relationships (e.g., joint and individual lectures and academic spin-offs). In this regard, Gibb (2000, p. 201) points out that “in such informal interaction lies the seeds of innovation”.
When considering the single mode that universities and M&SMEs adopt to manage collaboration, a large number of specific alternatives were proposed in the extant literature. In this respect, Apa et al. (2021) proposed the most complete list, which comprises 28 different alternatives, grouped according to the degree of formalization.
Breakdown of collaboration modes by content.
Source: own elaboration.
The Why issue
The Why issue analyzes three different elements, namely drivers inducing collaboration, enabling factors and barriers, which either facilitate or hinder the cooperation between universities and M&SMEs. In the following subsections findings related to these three elements will be discussed.
Drivers
Classification of drivers related to the university’s side.
Source: own elaboration.
Classification of drivers related to the M&SMEs side.
Source: own elaboration.
Comparing universities’ and M&SMEs’ drivers, it is worth noting they are quite different. This finding could represent an obstacle to the implementation of an effective and efficient collaboration between the two partners, as will be discussed in the subsection devoted to the emerged barriers. Moreover, drivers boosting universities to collaborate are mainly related to the more “traditional” missions (namely, education and research) than the so-called “third mission” (e.g., willingness to contribute to local development; Compagnucci and Spigarelli, 2020). On the other hand, M&SMEs are mainly driven by the search for access to facilities, competitive advantage, solving problems and acquiring knowledge.
Enabling factors
Classification of enabling factors.
Source: own elaboration.
As far as the context categories are concerned, the external environment was considered relevant only for the M&SMEs (e.g., industry technological characteristics). At the same time, the most investigated category in terms of preconditions is the one regarding cultural and managerial issues (e.g., M&SME’s ability to understand its technological needs vs comprehension of M&SME’s needs). In this respect, it is worth noting that economic and financial enabling factors (e.g., availability of tax incentives) are considered less relevant than the other preconditions. In addition, when considering the collaboration phases, the large amount of enabling factors are regarding implementation (e.g., university’s ability to codify scientific knowledge vs M&SME’s availability of data and information) followed by the search phase (e.g., university ability to select M&SMEs truly willing to innovate vs ability of the M&SME’s network to select the focus university). In contrast, the internalization phase was less investigated in terms of enabling factors; in this respect it is worth noting that only the absorptive capacity was frequently cited as a helpful item. This result is in line with previous analyses (Cohen et al., 2025) which demonstrate that when funding collaborative projects, governments typically prioritize companies with high absorptive capacity, emphasizing the importance of this factor for the success of U-ICs (Cui et al., 2023). Finally, the previous collaboration between partners was supported by 24 enabling factors (from the more previous personal relationships between academics and firm’s entrepreneurs/managers to the previous lecturing experience of M&SMEs’ representatives). However, 12 of these enabling factors are related to the M&SMEs’ side, while nine are shared ones. The significance of prior experiences is a crucial enabling factor strongly recommended by scholars (López-Martinez et al., 1994) for the effectiveness of cooperation agreements between universities and industry.
When considering the single enabling factor most cited for each partner, it emerges that the presence of focused organizational units (e.g., TTO and career offices) is the most critical on the university side, while – as pointed out earlier – the absorptive capacity is on the firm’s side. Finally, common trust and personal relationships are the most cited enabling factors among the shared ones.
Barriers
Classification of barriers.
Source: own elaboration.
As noted earlier for the enabling factors, the external context appears as more relevant as a source of barriers (e.g., M&SME’s location in a rural area). At the same time, the cultural and managerial preconditions (university lack of focus on knowledge transfer vs M&SMEs’ lack of perception of collaboration needs) emerge as the more relevant. In contrast, economic and financial barriers are less cited. Finally, the implementation phase emerges again as the most investigated one; in this respect, orientation to perfection and the lack of practical perspectives were cited as the most relevant for the university side, while lack of patience in conducting research projects was the most relevant barrier on the M&SME’s side.
In relation to the single barrier most cited for each partner, it emerges that university bureaucracy is perceived as the most critical issue hindering cooperation, together with the M&SME’s lack of absorptive capacity. This latter point constitutes a significant distinction between large companies and M&SMEs, since the former possess a greater capacity to absorb knowledge generated by universities due to their higher levels of human capital (Merritt, 2015). Finally, conflicting aims (publications vs products) emerges as the most relevant shared barrier.
Comparing evidence related to enabling factors and barriers, a certain level of reciprocity clearly emerges; in other words, several of the selected factors supporting the collaboration may reduce – and even overcome – the barriers found in the extant literature. For instance, the development of mutual trust (e.g., based on previous collaborations) may counterbalance the M&SME’s lack of confidence in universities’ capabilities and the university’s reluctance to deal with local small businesses.
Best practices breakdown by collaboration phase.
Source: own elaboration.
Finally, even if enabling factors and best practices were addressed by some of the selected documents, very few scholars paid specific attention to the evaluation of the collaboration performances. More specifically, 27 documents investigate such an issue but only 11 of them offer quantitative data regarding the collaboration projects’ success.
Based on the findings discussed earlier, a set of future research avenues will be proposed in the next section.
Research gaps and future research avenues
Over the period analyzed, the literature on U-ICs has not only expanded in volume but also achieved several significant advancements. The initial development concerns the conceptual foundations of the field. While early studies primarily concentrated on technology transfer and commercialization, subsequent contributions introduced more comprehensive frameworks such as the triple helix model (Etzkowitz and Leydesdorff, 2000) and open innovation (Chesbrough, 2003). These perspectives broadened the scope of analysis to encompass systemic and network-based forms of collaboration (Table 3). Although Bengoa et al. (2021) argue that U-ICs can contribute to sustainable development, our analysis did not identify explicit references to this issue within the selected papers.
A second advancement relates to the inclusiveness of actors. Although U-ICs have been studied since the late 1960s (Bengoa et al., 2021), M&SMEs only began to receive attention from the late 1980s, when scholars recognized both their liabilities of smallness and their agility in leveraging external knowledge. Nevertheless, substantial interest in M&SMEs as distinct actors emerged only in the early 2000s (Figure 1).
A third advancement concerns the type of collaboration between universities and M&SMEs. More specifically, while earlier studies focused predominantly on unidirectional technology transfer, more recent contributions have analyzed joint collaborations (Table 8). This evolution highlights a conceptual shift in the role of M&SMEs, increasingly framed not merely as passive recipients of knowledge but as strategic partners in research and innovation.
A fourth advancement pertains to the type of innovation fostered through U-ICs. Referring to the Oslo Manual taxonomy (OECD/Eurostat, 2018), our analysis shows that product and process innovations have been investigated since the late 1980s, whereas marketing and organizational innovations have been considered only since the early 2000s. This gap may be partly explained by the relatively lower scholarly attention devoted to these types of innovations, despite their specific relevance for M&SMEs (Freshwater et al., 2019).
The results presented earlier provide important insights regarding the ways in which the investigated collaboration takes shape. Therefore, this study offers valuable suggestions for scholars in terms of research gaps and directions for further research and methodological recommendations.
As far as methodological issues are concerned, only seven selected journal articles offer a specific definition of the investigated phenomenon. This is consistent with previous findings by Bengoa et al. (2021) concerning the broader literature on technology transfer. At the same time, Ranga et al. (2008) point out that M&SMEs are a quite a heterogeneous category; therefore, the concepts of “technology transfer” and “knowledge transfer” need to be more precisely defined based on the specific unit of analysis. Consequently, we suggest that future research should clearly define the conceptualization of U-IC they refer to, taking into account the specificities of the investigated M&SMEs.
Moreover, the findings presented above show that only 33 out of the 82 sampled studies are based on a single theoretical perspective and very few on more than one. This may result in being a limitation in new knowledge development. For instance, integrating the absorptive capacity lens (Cohen and Levinthal, 1990) with the Triple helix approach (Etzkowitz and Leydesdorff, 2000) could provide a robust theoretical foundation for generating new knowledge that would also be valuable for policy makers. This, in turn, could activate a spillover effect, which would enable further collaborations between universities and M&SMEs. In this respect, Ranga et al. (2008) and Jones and Corral De Zubielqui (2017) suggest that publicizing the success of relationships between the investigated partners may stimulate the development of further collaborations. In this vein, future research should be rooted in a single theoretical perspective or, even better, in a combination of them.
Summary of research gaps and Future Research Avenues (FRA).
Source: own elaboration.
The Who Issue
Findings clearly show the large majority of selected documents analyze universities and M&SMEs as general categories (see Tables 5 and 6), although some scholars have investigated specific typologies of these actors (e.g., public vs. private university; hi-tech vs. low tech companies). However, no previous studies offer a comparison between different actors. For instance, universities including engineering departments (as in the case of Capaldo et al., 2016) could be compared with Medical Research University (Dahlborg et al., 2017) and/or Business School (Dabic et al., 2016). This approach is consistent with the assumption that there does ‘not exist a single mode or template that fits for all universities’ (Grimaldi et al., 2021: p. 863). At the same time, the experience of young firms (Bellini et al., 2019) could be compared with the older ones (Garcia-Perez-de-Lema et al., 2017). Therefore, we put forward the following future research avenue (FRA). FRA 1: Future research should investigate the different types of involved actors (University and M&SMEs) and also compare them.
With a specific focus on the university side, the findings clearly indicate that only TTOs were analyzed by different scholars (11 documents in total), while other organizational units (e.g., university research and innovation units) and/or actors (e.g., university faculty) were rarely addressed (Table 5). In this vein, Bengoa et al. (2021) affirm that technology transfer mechanisms other than TTOs deserve specific attention. For instance, universities hosting different departments (e.g., business vs engineering) may interact in a diverse way with M&SMEs. Therefore, we suggest the following FRA: FRA 2: Future research should further investigate the single university components interacting with M&SMEs, also analyzing their involvement in the different steps of the collaboration.
On the other hand, given that M&SMEs represent a highly heterogeneous category (Ranga et al., 2008), we propose the following FRA: FRA 3: Future research should be implemented by adopting a comparative approach in terms of firms’ size and industry.
The What issue
Regarding the What issue in terms of collaboration outputs, it emerged that all the four innovation typologies of the so-called “Oslo Manual” (OECD/Eurostat, 2018) were investigated (Table 10). However, there is a lack of adequate analysis regarding differences (if any) in how the collaboration occurs based on the expected outputs in terms of innovation typology. For instance, the fear of intellectual property (IP) rights disclosure (emerged as one of the barriers on the firm’s side) seems to be more relevant in the case of product and/or process innovation, while easiness to copy innovations could be more critical in the case of organizational and marketing innovation. Therefore, we propose the following FRA: FRA 4: Future research should specifically examine whether the expected type of innovation influences the interactions between the involved actors at various stages of the collaboration.
The When issue
When addressing the When issue, it emerged that longitudinal studies with an extended time span are very rare, particularly those covering the longest durations (at least 5 years). In this context, it is worth noting that innovations in legislation (e.g., university evaluation systems, including also the so-called “third mission”) and/or in technology (e.g., digital ones) could impact the ways in which Universities and M&SMEs interact (Adomako and Nguyen, 2023). At the same time, the collaboration could be affected by changes at the single actor level, as in the case of the establishment of TTO (on the University side) or entrepreneurial succession (on the M&SME side). Therefore, we formulate the following FRA: FRA 5: Future research should adopt a longitudinal approach in order to verify any evolution in terms of factors affecting the U-ICs in general and/or the single actor.
The Where issue
As pointed out earlier, only a few of the sampled articles offer a comparison of the phenomenon among countries. More specifically, three alternative comparative approaches were implemented: (a) different countries (e.g., Biro, 2015); (b) different regions within the same country (e.g., Amano-Ito, 2020; Capaldo et al., 2016); and (c) regions located in different countries having similar features (e.g., in terms of the Regional Innovation Scoreboard features). In this respect, Garcia-Alvarez-Coque et al. (2019) pointed out that universities relate differently to external actors on the basis of the environment in which they are embedded. Moreover, Grimaldi et al. (2021) highlighted that university participation in the local economy is affected both by regional and national policies and legislation. Finally, it must be noted that the observed predominance of studies conducted in high-income countries (Table 12) may obscure important regional dynamics within those nations, while leaving unexplored the specificities of emerging or peripheral regions. Consequently, we propose the following FRA. FRA 6: Future research should aim to compare the investigated phenomenon across different geographical contexts.
The How issue
The conducted analysis identified 23 different collaboration modes, including collaborative research and informal relationships. However, no previous studies have investigated the relationship (if any) between firms’ characteristics (in terms of size, geographic location and innovation intensity) and the adopted collaboration mode. As an example, co-authorship is primarily expected in companies employing human resources with a PhD; at the same time, MSc students transfer could be easily activated also in the case of micro and small companies. Therefore, we suggest the following FRA: FRA 7: Future research should verify if the firm’s characteristics influenced the choice of collaboration mode.
The Why issue
Pereira and Franco (2023) pointed out that scholars generally emphasize the barriers that hinder the collaboration between universities and M&SMEs. However, specific knowledge of such barriers is still in its infancy. Moreover, it seems there is no adequate debate on how to overcome them. Therefore, we suggest the following FRA: FRA 8: Future research should implement a holistic approach which simultaneously considers enabling factors and barriers, also considering the firms’ features (e.g., size, industry and innovation propensity).
Such a holistic approach emerges as critical during the internalization phase, when the presence or absence of absorptive capacity within the M&SMEs may represent either a barrier or an enabling factor (e.g., Apa et al., 2021; Rajalo and Vadi, 2021). Moreover, findings in previous studies are not always definitive. Finally, as pointed out by Hervas-Oliver et al. (2012), absorptive capacity may be assessed based on different variables, not limited to those related to R&D. Consequently, we propose the following FRA: FRA 9: Future research should further investigate the role of presence or absence of absorptive capacity in the M&SMEs, taking into account their specific characteristics (e.g. size, geographic location, industry). Moreover, special attention should be given to the variables adopted to assess this capacity.
Finally, very few scholars have paid specific attention to the evaluation of collaboration performance: while 27 documents investigated this issue, only 11 provided quantitative data. This issue is quite relevant, especially in terms of so-called “societal impacts”, i.e. any “positive or negative change originated in the U-IC context that can directly or indirectly affect individuals, organizations, communities and society in general” (Cohen et al., 2025: p. 346). Additionally, very few documents contain suggestions regarding best practices for enhancing the effectiveness of the search and implementation phase. Therefore, we suggest the following FRA. FRA 10: Future research should further investigate the degree of success of the collaboration and identify best practices for its different phases.
Concluding remarks
This study is aimed at shedding new light on the collaboration between M&SMEs and universities. This research aim was reached through an SLR of 82 Scopus and or WoS indexed articles published between 1987 and 2023. Collected data were analyzed according to the 5 Ws&1H approach referring to the following perspectives: (1) who (involved actors); (2) what (collaboration directionality, inputs and outputs); (3) when (investigated time span); (4) where (analyzed countries); (5) how (collaboration modes) and (6) why (drivers, enabling factor and barriers).
An initial contribution of this article is the comprehensive interpretive framework adopted to explore the 5 Ws&1H dimensions (Figure 2). This methodology allowed us to identify very heterogeneous positions among scholars, with the proposition of several different variables influencing collaboration, especially in the case of enabling factors and barriers. It also highlighted the lack of a common conceptualization of the phenomenon under study, combined with a relatively scarce adoption of theoretical perspectives. On the other hand, a wide range of research methodologies was used, including both quantitative and qualitative approaches. Due to the fragmentation of empirical evidence and the lack of a common perspective, a further contribution on the academic side is represented by the 10 FRAs proposed to support future research efforts.
When compared with the most recent review in the field, namely Pereira and Franco (2022), our study provides several additional insights. These authors conducted a bibliometric analysis of articles published between 1995 and 2019; therefore, their focus was mainly on publication patterns, citation counts, and the general characterization of SME–university collaborations. In contrast, our review extends the time frame from 1987 to 2023 and adopts an SLR review combined with a 5 Ws&1H interpretative framework. This allows for a more nuanced exploration of the phenomenon, capturing not only who is involved and how collaborations are implemented, but also why they succeed or fail. Moreover, our findings highlight the absence of a shared conceptualization, the limited use of theoretical perspectives, and the heterogeneous role of enablers and barriers, aspects that were not explicitly addressed in the earlier review. In doing so, our study complements and expands on Pereira and Franco’s (2022) contribution, offering both a broader empirical base and a deeper theoretical contextualization of M&SME–university collaborations.
To sum up, our extant literature review confirms that collaboration with universities may represent a win-win relationship (Pujotomo et al., 2023) also for M&SMEs. However, adequate efforts should be implemented to cope with the large number of barriers that may hinder this relationship. Therefore, enabling factors and best practices should be adequately taken into account by all the involved actors. In this respect, our findings offer useful insights for university managers (especially those involved in technology transfer activities), M&SMEs entrepreneurs and managers, and policy makers. Among the implications that emerge for the former two categories (university managers and M&SMES governance), the most important is regarding the relevance of cultural preconditions (e.g., Bjerregaard, 2010; Rajalo and Vadi, 2021), often due to the inadequate perception of the counterpart perspectives in terms of aims (e.g., publication vs new product and process) and owned competences. Therefore, networking activities should be intensively promoted not only by universities but also by entrepreneurs’ association, in order to develop mutual knowledge, appreciation and trust (Grant et al., 1996). At the same time, universities should carefully select academics who are really interested in collaborating with M&SMEs, evaluating not only their technical and technological capabilities but also their relational skills (Bjerregard, 2010; Thompson and Homer, 2005).
In this regard, it is crucial to acknowledge that, generally, individual academics are the primary drivers of these collaborations, rather than the university as a singular institution (Taxt, 2024). Historically, the principal priorities of universities have been the quantity of high-impact publications and patents, the volume of research funding secured, the perceived quality of academic programs offered, and the employability prospects of their graduates. These priorities may not be entirely congruent with collaborations involving M&SMEs, given their comparatively limited impact on research and educational objectives when contrasted with engagements with large corporations. However, the so-called third mission – conceptualized as encompassing both collaboration with companies and other public administrations, and societal impact – has increasingly emerged as a pertinent issue within universities’ strategic frameworks (Wilson et al., 2024). Consequently, the promotion of collaboration with M&SMEs should be progressively integrated into universities’ institutional objectives. Concurrently, university administrators should cultivate the capacity to meticulously select M&SMEs demonstrating a significant willingness to engage (Caputo et al., 2002) and to identify companies with technological requirements aligned with those previously addressed by the university academic staff (Lavery and Stratford, 2003). In this manner, these prior collaborations can serve as valuable references. Simultaneously, M&SMEs are also advised to implement specific tools to identify universities that best align with their technological needs (Ran et al., 2020). In this respect, it is noteworthy that collaboration with universities can offer a dual advantage for M&SMEs, as this factor is considered by large companies when selecting their suppliers (Gupta and Barua, 2018).
Finally, universities should integrate the so-called “third mission” (which includes technology transfer activities) with the education mission developing Masters courses specifically devoted to the technological transfer, particularly for M&SMEs (Ilkan et al., 2010; Morales et al., 2018). In this respect, Besednjak Valič et al. (2023) explicitly request policy makers to support curricula adjustments to support the demand for skilled human resources.
As far as policy makers are concerned, it clearly emerged that economic and financial issues are of relatively lesser importance. However, financial aids become of primary importance in the case of programs involving Master of Science students supported by academics. Such exchange programs – which require the availability of adequate scholarships for the involved students - were cited as an effective first step of a long-term collaboration according to an incremental approach (Caputo et al., 2002). At the same time, policy makers have a critical role in facilitating the initial contact between M&SMEs and universities and creating mutual trust between them. Finally, governments should assure a stable legal environment, including the legislation issues (Besednjak Valič et al., 2023).
Beyond these implications, our findings also suggest further considerations for practitioners. For M&SME managers and entrepreneurs, a key priority is to develop organizational routines that enhance their companies’ absorptive capacity, enabling them to effectively internalize and apply external knowledge. This implies investing not only in technical preparedness but also in relational competences, which are crucial to sustain long-term partnerships with academic counterparts. University managers, on their side, should recognize that collaborations with smaller firms require different approaches compared to the ones with large corporations, including more flexible contractual solutions, stronger trust-building mechanisms, and continuous alignment of objectives. Dedicated liaison offices or specialized units for M&SME engagement could significantly reduce transaction costs and increase collaboration effectiveness.
Like all structured literature reviews, this study is subject to several methodological and conceptual limitations. First, the analysis was restricted to articles published in English and indexed in the Scopus or WoS dataset, which may have excluded relevant contributions published in other languages, in non-indexed journals, or in the grey literature. Second, although the systematic search strategy was carefully designed, publication bias and database coverage may still have influenced the final sample. Future research could therefore extend the scope to include indexed conference papers and book chapters, as well as other documents listed in Google Scholar, regardless of language. Third, the interpretative framework adopted—the 5 Ws&1H approach—enabled a comprehensive categorization of findings but inevitably reflects the authors’ analytical choices, which may not capture all possible dimensions of the phenomenon. Finally, our SLR focused primarily on the content of the sampled documents without addressing bibliometric aspects. In this respect, future research could explore the intellectual structure of the field. A citation and co-citation analysis could shed light on whether the literature on U-IC with M&SMEs is evolving into a coherent disciplinary community, in the sense of an “invisible college” (Crane, 1972), or whether it remains a dispersed set of contributions only loosely connected by shared keywords. Such an investigation would also make it possible to assess the extent to which a common conceptualization of U-ICs exists. In this regard, the use of knowledge-mapping techniques—for instance, graph theory and AI-based bibliometric tools—could provide valuable insights into the patterns of connectivity among authors, journals, and concepts, thereby clarifying the intellectual foundations and future trajectory of the field.
Future reviews could build on these limitations in several ways. Periodic updates would ensure that new publications are systematically integrated, while bibliometric approaches (e.g., citation and co-citation analyses) could help uncover the intellectual structure of the domain. Moreover, meta-analytical techniques could be employed to quantify effect sizes and identify robust empirical patterns across studies. Such complementary approaches would not only overcome some of the limitations of the present study but also contribute to developing a more cumulative and theoretically grounded understanding of U-IC involving M&SMEs.
Footnotes
Funding
This work has been funded by the European Union - NextGenerationEU under the Italian Ministry of University and Research (MUR) National Innovation Ecosystem grant ECS00000041 - VITALITY - CUP E13C22001060006. Project title: “VITALITY - Innovation, digitalisation and sustainability for the diffused economy in Central Italy (Ecosistema di Innovazione, Digitalizzazione e Sostenibilità per l'Economia Diffusa nell'Italia Centrale)”
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.