Open innovation in science: assessing the formation and function of SME-university collaborations through the proximity matrix

ABSTRACT

As university-industry collaboration is regarded as an important practice within the open innovation in Science (OIS) framework, this paper assesses the formation and function of these collaborations using the ‘proximity matrix’, evaluating similarities between actors through evaluating their closeness in terms of distance, network membership, knowledge base and working practices. Through presenting analysis of 22 in-depth interviews with SMEs, the findings outline that the process of SME-university collaboration is driven by the ability of the firms to both access (through social proximity) and understand (through technological and organisational proximity) their university partners. Furthermore, the results also suggest distinct roles for each proximity.

KEYWORDS:

• University-industry links
• SMEs
• proximity matrix
• open innovation in science

1. Introduction

The practice of Open Innovation in Science (OIS), defined as ‘purposively enabling, initiating, and managing inbound, outbound, and coupled knowledge flows and (inter/transdisciplinary) collaboration across organisational and disciplinary boundaries along all stages of the scientific research process, from the conceptualization, to exploration and/or testing, and documentation’ (Beck et al. pg., 3), sees alliances between universities and industry as an important conduit for knowledge transfer. SMEs in particular are viewed as benefitting from OIS as they may lack the resources to innovate alone, therefore providing them with access to knowledge from scientific research undertaken within universities (Hewitt-Dundas and Roper 2017; Vicente-Saez and Martinez-Fuentes 2018). Indeed, there is a substantial evidence that university collaborations have a positive impact on SMEs through promoting organisational learning, market development, new business opportunities, and innovation (Dada and Fogg 2016; De Zubielqui, Jones, and Lester 2016; Rosli et al. 2018; Lauvås and Steinmo 2019).

Importantly, SME-university collaboration is increasingly perceived to be a socio-technical process, enabled by similarities between the actors in terms of relational and cognitive factors which promotes the smooth transfer of knowledge across organisational boundaries (Steinmo 2015; AL-Tabbaa and Ankrah 2016; O’Reilly and Cunningham 2017; Rajalo and Vadi 2017; Al‐Tabbaa and Ankrah 2019). Therefore, these similarities in essence constitute the boundary conditions facilitating the formation and function of SME-university collaborations through reducing any potential barriers to effective cooperation that may exist to be minimised, the so called ‘two worlds paradox,’ (Hewitt-Dundas, Gkypali, and Roper 2019; Beck et al. 2020). Given the importance of ‘boundary conditions’ to the successful application of OIS practices (Beck et al. 2020) and individual attributes that facilitate these collaborations (Perkmann et al. 2013; Johnston and Huggins 2018), a clear understanding of the interplay of these factors that draws connections between similarities amongst actors and specific activities that either enable or constrain SME-university collaboration is required.

In order to assess how these relational and cognitive factors may provide the effective boundary conditions to enable SME-university collaborations to both form and function, this paper uses a ‘proximity matrix’ framework, which is designed to capture similarities between actors through evaluating the closeness of partners (Johnston and Huggins 2021). The proximity matrix captures similarities between actors in terms of their spatial proximity (physical distance), social proximity (network membership), organisational proximity (working culture and methods), and technological proximity (knowledge base) (Knoben and Oerlemans 2006; Aguilera, Lethiais, and Rallet 2012; Balland, Boschma, and Frenken 2015). This framework, therefore, draws on insights from the geography of innovation, inter-organisational alliances, and networking literatures to build a unifying lens through which the similarities of SMEs and their university partners can be examined. Given current calls for a better understanding of the micro-level practices, boundary conditions, and individual characteristics that influence SME-university collaboration (Albats, Fiegenbaum, and Cunningham 2018; Cunningham and O’Reilly 2018; Beck et al. 2020), this paper adds to the understanding of the formation and function of SME-university collaboration by addressing two research questions: 1) To what extent does each proximity influence both the formation and function of SME-university links? 2) At which stage of the project does each proximity influence the collaboration process between SMEs and universities?

In order to address these questions, the paper draws on in-depth interviews with 22 UK SMEs which had engaged in formal collaborations with universities through the Knowledge Transfer Partnership (KTP) programme. The resulting thematic analysis examined the roles of proximities in relation to the formation of the collaborative link, i.e. how it came into existence, and secondly, the functioning of the link, i.e. the processes that underpinned the workings of the project. The paper’s findings contribute to a deeper understanding of OIS through SME-university collaboration by suggesting that boundary conditions necessary require SMEs to be able to both access (through social proximity) and understand (through technological and organisational proximity) their university partner. Furthermore, the results also suggest distinct roles for each proximity; for example, the influence of social proximity is observed in the formation of the collaborations, while the influence of both technological and organisational proximity is observed in both the formation and function of the collaborations. In addition, there is evidence that technological and organisational proximity may evolve during the projects. Finally, the results do not suggest the primacy of any type of proximity in the formation and function of SME-university links, justifying the use of the proximity matrix.

The paper is structured as follows: Section 2 outlines the conceptual and theoretical background. Section 3 presents the methodology. Section 4 presents the findings and a discussion of their implications. Finally, Section 5 concludes.

2. Theoretical and conceptual background

2.1. Open innovation and SME-industry collaboration

Within the Open Innovation in Science (OIS) paradigm, university-industry collaboration is highlighted as a key mechanism through which this practice may occur (Beck et al. 2020). OIS provides ‘a unifying foundation for advancing our understanding of antecedents, contingencies, and consequences related to applying open and collaborative research practices along the entire process of generating and disseminating new scientific insights and translating them into innovation.’ (Beck et al. 2020, 4).

Open and collaborative innovation practices are regarded as especially important for SMEs to improve their innovativeness by allowing them to circumvent their resource constraints (Ebersberger et al. 2012). In particular, university collaboration is seen as an important means by which open innovation can be achieved by SMEs (Hewitt-Dundas and Roper 2017; Vicente-Saez and Martinez-Fuentes 2018). despite the importance of OIS through university collaboration, in an SME context these linkages are often overlooked. This perception is perpetuated by evidence to suggest that smaller firms are less likely to collaborate with universities than their larger counterparts (Bodas Freitas et al. 2006; Fontana, Geuna, and Matt 2006). However, this does not mean that SMEs do not engage with universities, only that vis a vis larger firms, it is the latter that are more prone to developing these links. Furthermore, the relationship between firm size and university collaboration is more nuanced, with SMEs tending to engage in a higher number of projects than larger firms (Motohashi 2005) while also being less likely to engage in formal interaction with universities than larger firms (Bodas Freitas et al. 2013). SMEs may also focus on longer term projects, centred on organisational learning that is less hurried and more deliberate in nature (Broström 2010). Importantly, the advantages to firms from engaging with a university, such as improving understanding, gaining knowledge, problem solving, and training the workforce appear to be unrelated to firm size (Bishop, D’Este, and Neely 2011), while smaller firms have been found to benefit higher levels of growth after receiving public funding for research projects (Vanino, Roper, and Becker 2019). Accordingly, SMEs do benefit from university collaboration as it promotes organisational learning, new business opportunities, and innovation (Dada and Fogg 2016; De Zubielqui, Jones, and Lester 2016; Rosli et al. 2018; Lauvås and Steinmo 2019). Given these findings, universities are regarded as important sources of external knowledge for SMEs (Johnston and Huggins 2021).

However, promoting OIS though SME-university collaborations is not always a seamless process as translating science into practice across organisational boundaries may involve overcoming barriers between the actors involved. These barriers are typically manifested as differences in the organisational contexts and institutional environments of the project partners, with universities typically focussing on exploration activities and SMEs on exploitation activities (Lavie and Drori 2012; Messeni Petruzzelli and Rotolo 2015). In particular, there may exist differences in methods of working resulting from different motivations, with SMEs focussed on commercialisation whereas academics concentrate on pursuing novel research (D’Este and Perkmann 2010; Garman 2011; Lam 2011) SMEs and universities may also work to differing time scales, with the former operating to a stricter timescale dictated by commercial demands (Walsh et al. 2007). These differences are summed up by ‘the two-worlds paradox’, which suggests that these differing contexts and motivations mean that a degree of incommensurability exists between the parties (Hewitt-Dundas, Gkypali, and Roper 2019).

Given these barriers to operationalising OIS and transferring knowledge across organisational boundaries, SMEs place great significance on seeking academic partners with the appropriate knowledge and expertise (Mäkimattila, Junell, and Rantala 2015; Johnston and Huggins 2018). For SMEs, the partner selection process focuses on ensuring that their partner is credible in terms of being able to delivering the knowledge and expertise they require (Johnston and Huggins 2018). To achieve this, ‘short institutional distances’, are required between the SME and the university partner (Bjerregaard 2009), referring to similarities between the partners in terms of social, technical, and organisational backgrounds. Indeed, these have been identified as the defining features of SME-university collaborations characterised by effective communication and a successful outcome (Rajalo and Vadi 2017). Similarly, Steinmo and Rasmussen (2018) found that the ability of actors to relate and understand one another, or possessing sufficient ‘cognitive social capital’, enables the development of trusting relations and the development of a shared goal between SMEs and their university partner.

Therefore, relational and cognitive similarities between the actors is regarded as an important element in the formation and function of SME-industry links. Accordingly, the following section sets out a framework to assess these similarities based around a matrix of proximities, or the closeness of actors, spatially, socially, organisationally, and technologically.

2.2. Proximities and SME-university collaboration

Proximities capture the closeness of partners across several dimensions including physical distance or location (spatial proximity), network membership (social proximity), cognitive understanding (technological proximity), and similarity or working culture (organisational proximity) (Boschma 2005; Aguilera, Lethiais, and Rallet 2012; Balland, Belso-Martínez, and Morrison 2016). The spatial proximity of partners, i.e. their physical closeness, is typically regarded as a positive factor in their formation (D’Este, Guy, and Iammarino 2013; Giuliani and Arza 2009a; Johnston and Huggins 2016a; Laursen, Reichstein, and Salter 2011). This physical closeness allows the actions of potential partners to be observed, permitting an assessment of their effectiveness (Wood and Parr 2005; Gulati 2007). Furthermore, spatial proximity increases the intensity of collaborative links, through promoting greater levels of face-to-face interaction (Storper and Venables 2004). Thus, the spatial proximity of partners offers an opportunity to observe potential partners and assess their organisational routines (Capaldo and Petruzzelli 2014).

Similarly, social proximity is based on connections resulting from friendships, shared networks, and prior public and private interactions between actors (Hansen 2015). While actors may have divergent backgrounds and work in diverse industries, their interaction through events, meetings, fora, or broad based membership organisations may breed a familiarity between them that facilitates collaboration (Mattes 2012). As such, social proximity can enable collaborations through partner search and selection (Grimpe and Hussinger 2013).

The third type of proximity, technological proximity, refers to similarities of knowledge, expertise, experiences, know-how of particular processes, machinery, or tools between partners (Knoben and Oerlemans 2006). This is viewed as having a positive effect on the collaborative ties as it means all actors understand the fundamentals of a particular field (Nooteboom 1999). Where actors also share the same technical language they can more easily understand each other, making collaboration more effective (Huber 2012).

Organisational proximity has been conceptualised in terms of the similarities between agents based on shared knowledge, methods of working, relationships, and culture and ‘belong to the same space of relations’ (Knoben and Oerlemans 2006; Aguilera, Lethiais, and Rallet 2012; Kuttim 2016). Within the extant literature, this has been conceptualised as prior experiences of working together which allows actors to learn how one another works in terms of culture and methods. Consequently, organisational proximity has be found to have a positive effect on the formation of U-I links (Aguilera, Lethiais, and Rallet 2012; D’Este, Guy, and Iammarino 2013).

It is also suggested that these proximities may be dynamic in nature, (Balland, Boschma, and Frenken 2015), evolving in the course of collaborations as actors learn to work together effectively (Bjerregaard 2010; Steinmo and Rasmussen 2018). Given that proximities are considered to be dynamic in nature (Balland, Boschma, and Frenken 2015), it also makes sense to consider them as continuous rather than binary. Accordingly, proximities should not have to be treated as dichotomous, either present or otherwise, instead they can be assessed in terms of their degrees of completeness (Johnston and Huggins 2021).

However, this focus on proximities has been criticised for being overly simplistic as innovation may result from complex patterns and interactions of proximities, whereby complementarities exist but also substitution effects are observed (Mattes 2012; D’Este, Guy, and Iammarino 2013; Johnston and Huggins 2016a). In addition, there may be drawbacks associated with each type of proximity. For example, spatial proximity may not convey any advantages other than convenience (Presutti et al. 2017) or may just be a substitute for other proximities (Mattes 2012). Social proximity may mean that actors simply share the same connections, therefore accessing the same knowledge (Maurer and Ebers 2006; Broekel and Boschma 2012). Similarly, technological proximity may also indicate the heterogeneity of knowledge (Nooteboom 1999). Finally, organisational proximity may hinder firm flexibility through locking it into an unproductive relationship (Boschma 2005). Therefore, given this critical appraisal of the role of proximities in the university-industry collaboration, there is a need to understand exactly how they influence the formation and function of SME-university collaborative links.

3. Methodology

To explore the experience of SMEs in the process of OIS through university collaboration this paper utilises data from twenty-two in depth interviews of SMEs participating in formal collaborative projects with universities via the Knowledge Transfer Partnership (KTP) programme in the UK. KTPs are an important source of funding for supporting industrial R&D in the UK, designed to assist businesses with innovation to promote their growth and competitiveness. The aim of the programme is to promote collaboration between firms and universities, ensuring that the ‘latest academic thinking’ is introduced into the firm to promote innovation. These projects can last between one and three years and are part-funded by a public grant to cover the costs of the project. Public funding for KTPs covers up to 67% of the project’s costs for SMEs, with the remainder funded by the business involved. In return, the business is supported by an academic supervisor from the partner university as well as a full time ‘Associate’, typically a graduate in a relevant discipline, who carries out the day-to-day operations of the project. Indeed, the success of the programme is highlighted by the fact that applications to the scheme has been increasing over the last few years (Johnston and Huggins 2021)

Details of firms participating in the KTP programme are available through the Gateway to Research website, a freely available resource that provides details of all publicly funded research projects in the UK. SMEs involved in the KTP programme were identified using this resource, the main criterion being that they had embarked on and completed a KTP within the previous 5 years (2013–2018). This ensured that the collaboration remained relatively fresh in the firm’s organisational memory while also allowing sufficient time for the project to have been completed and the firm to have an idea of the impacts of the project. All SMEs meeting this criterion (54) were contacted and 22 agreed to be interviewed, a response rate of 42.3%. The analysis therefore focuses on a diverse range of SMEs from sectors such as software, finance, biotechnology, consultants, manufacturing, and marketing, varying in terms of size from 3 to 120 employees, and turnover from £130,000 pa to £35,000,000 pa (See Table 1 for details) reflecting the heterogeneity of the SME sector (Branzei and Vertinsky 2006; Franco and Haase 2015).

Table 1. Details of participating SMEs

Alias	Sector	Partner	Employees	Turnover	Age
Arc-Tech	Architectural Consultants	Coventry University	50	£3 million	5
Black-Electronics	Electronics Manufacturing	Sheffield Hallam University	27	£4 million	31
Blue Finance	Accident investigation software and consultancy	University of the West of Scotland	15	£4 million	22
Blue-Design	Design Consultancy/3D Printing	Manchester Metropolitan University	6	£300,000	8
Brown Media	Media consultants	Salford University	31	£800,000	9
Cyan-Soft	IT Support	London Metropolitan University	17	£800,000	3
Data-Tech	Financial data software/analysis	Bournemouth University	12	£500,000	17
Edu-Tech	Education consultant/online assessment	University of Wolverhampton	12	£185,000	23
Gen-Tech	Biotechnology – DNA Therapies/technology	Birmingham University	12	£9 million	4
Green-Soft	Suppliers of analytical instrumentation/equipment	Anglia Ruskin University	7	£1.2 million	6
Indigo-Consultants	Health and Social Care Consultancy	Brunel University	7	£130,000	10
Magenta-Design	Design of play areas	Sheffield Hallam University	40	£4 million	10
Magnolia-Support	Project management/support	University of Exeter	50	£13 Million	5
Marine Consultants	Oil and gas consultancy	University of Exeter	40	£7-8 Million	9
Mine-Tech	Scientific analysis of geological material	Southampton University	20	£4.5 million	14
Motor-Tech	Electric motor simulation software	Glyndwr University	23	£1 million	9
Patient-Tech	Medical Consultancy	University of Westminster	3	£850,000	5
Pink-Soft	Software	Aston University	27	£3.5 million	13
Purple Media	Retail/online marketing consultancy	Durham University	120	£35 million	9
Red-Soft	Software	University of Plymouth	24	£230,000	10
White-Electronics	Electronic Manufacturing	University of Nottingham	80	£9-12 Million	16
Yellow-Soft	Marketing Solutions/Data Management	University of the West of England	65	£5 million	20

The methodological approach adopted in this study was designed to allow a comprehensive account of the firms’ experiences of this U-I collaboration, detailing the complexities and nuances of their participation. To achieve this, semi-structured interviews were undertaken with key personnel within collaborating SMEs, either the managing director or project director. All interviews were recorded and then fully transcribed afterwards; in conjunction, contemporaneous notes were also made in order to capture the nuances of the interview, which assisted with attaching meaning to the transcripts produced (Mishler 1986; Larty and Hamilton 2011).

The use of a narrative approach to data collection was influenced by the increasing use of these techniques in both the organisation studies (Czarniawska 1998; Boje 2001) and entrepreneurship literatures (Gartner 2010). This approach provides a useful technique for understanding the events occurring during U-I linkages, allowing the participants’ experiences to be captured. Consequently, the semi-structured interviews covered a broad range of topics, loosely following an outline based on Perkmann et al.’s (2013) factors that characterises the process of U-I collaboration:

• Project details. The focus of the project, its aims and objectives.

• Idea generation. To understand the both the genesis of the project and he motivation for engaging with a university.

• How the collaboration worked. Exploring the workings of the collaboration, including which personnel were involved, whether the partners were always able to communicate effectively, how the knowledge transferred into the firm was used, how the knowledge was applied, whether the participants learned from the process, and whether the collaboration was a mutual learning experience.

• Project outcomes. Finally, the outcomes of the project were discussed in terms of what resulted from the collaboration and whether the project aims were met.

The use of narratives in this context allowed the capture of first-person descriptions of events surrounding the experiences of SMEs participating in collaborative projects with universities from the personnel within the firms involved. A loose script was used to enable the use of a ‘life-story approach’ (Johansson 2004) which allowed the respondents to tell their stories and illustrate the events that occurred. Importantly, this approach allows the respondents to outline their experiences from their point of view, rather than based on a pre-defined set of factors (Clandinin and Connelly 2000). This approach allowed the development of ‘focussed-interview narratives’ (Mishler 1986), i.e. the building the narratives through exploring the respondents’ answers, but also allowing for the probing of respondents where greater detail was required, all the while following the rules of not interrupting and letting the respondent outline various events. As such, adhering to these protocols ensured that all respondents were given a voice (Bauer 1996), and enabled the creation of a ‘multi-voiced’ account of the projects in question (Fletcher 2007; Ericson 2010).

Subsequent analysis was concerned with connecting the stories into a plot in order to highlight the activities that occur within the firms (Steyaert 2007). Consequently, the analysis draws on the structuralist approaches to analysing narratives (Larty and Hamilton 2011). The analysis used NVIVO, a specialist software package designed for analysing qualitative data, to examine the transcripts. The first act was to read the transcripts firstly to develop the main plots (Boje 2001). These plots were two-fold; the first were based on the stage of the project to understand the process of the collaboration in the context of the project stage. Therefore, the transcripts were examined and the coded thematically in terms of the stage of the project: 1) the initial stage (pre-project); 2) the project stage (undertaking the project); and 3) the post-project stage (once the project was completed) (Ryan and Bernard 2003).

Second, the data was then re-examined on a plot-by-plot basis using the conceptual framework as a guide to identify the influence of proximities on the formation and function of these collaborations. This firstly involved the development of a narrative for each project from the transcript detailing the course of the project from idea generation through to its completion (Steinmo and Rasmussen 2018). These narratives were then analysed to identify the existence of proximities between the partners, assess their role in the formation and function of the projects, and then develop theoretical propositions around their roles in the process. Thus, the thematic analysis involved a two-stage process of analysis, the first stage coded the transcripts according to a pre-determined stage framework and the second involved a thematic analysis of the narratives to assess the roles of proximities (Fereday and Muir-Cochrane 2006). In order to achieve this, the proximity constructs were first of all defined according to the extant literature, providing a theoretical justification for the coding frame (Eisenhardt 1989).

Scholars have utilised both direct and indirect measures for capturing proximities (Fitjar, Huber, and Rodríguez-Pose 2016). Direct measures involve asking respondents to rate the importance of pre-set measures of proximities in terms of their perceived closeness of their partner in terms of factors such as knowledge base, working practices, or network membership (Fitjar, Huber, and Rodríguez-Pose 2016; Garcia et al. 2018; Nilsen and Lauvås 2018). In contrast, indirect measures focus on developing an account of the collaboration and then undertaking an ex-post evaluation that identifies proximities between actors through behaviours and establishes causality through assessing their impact on events. Therefore, the latter approach allows the entire spectrum of socio-relational factors to be examined in the context of the formation and function of SME-university collaborative links (Broström 2010; AL-Tabbaa and Ankrah 2016). In addition, given the heterogeneity of SMEs, their alliances and innovation practices (Branzei and Vertinsky 2006; Franco and Haase 2015), both variations and similarities in behaviours and events can be observed.

In this paper, an indirect approach is used to identify proximities, assessing their roles through examining the narratives derived from first-person accounts of the collaboration. Consequently, as the respondents were not asked about their perceptions of the closeness of their relationships with their collaborative partner, this had to be deduced from the actions and events that occurred. This allowed a broad definition of each proximity to be specified a priori and then examine the events associated with these to establish relationships between the observed proximities and subsequent events (Eisenhardt 1989)

However, proximities are not uncontested in nature, nor defined in a straightforward manner. For example, spatial proximity focuses on the physical closeness of the partners and can be operationalised through examining references to distances or the locales in which they operate. Social proximity was operationalised by identifying ties between the actors based on friendships, shared connections, and previous interactions through membership of communities or organisations or attending events (Hansen 2015). Technological proximity between the actors was operationalised through observing the SMEs’ perceptions of the similarities between and their understanding of their university partners’ knowledge and expertise of particular processes or technologies (Knoben and Oerlemans 2006). Finally, organisational proximity was operationalised where similarities and familiarities between agents were identified in terms of working with universities, and understanding their methods and cultures of working (Knoben and Oerlemans 2006; Aguilera, Lethiais, and Rallet 2012; Kuttim 2016).

The results were then tabulated to assess the relationships that existed through linking identified proximities with events and outcomes, thereby developing in-depth insights across all SMEs. This allowed the roles played by the various proximities to be identified and then assessed considering the theorised roles set out in the literature. These were examined firstly in relation to the formation of the collaborative link, i.e. how the tie came into existence and the processes that resulted in the collaborating partners working together. Secondly, the proximities were examined in terms of the function of the link, i.e. the processes that underpinned the workings of the project. Therefore, the process resembles the ‘systematic combining’ outlined by Dubois and Gadde (2002), whereby the theoretical is confronted with the empirical in order to juxtapose actual events with hypothetical predictions from theory.

4. Findings and discussion

4.1. Spatial proximity

Spatial proximity has a limited influence on the formation and function of SME-university collaborations. Spatial proximity is perhaps the simplest construct to observe as it relies on examining the physical distance between the actors. The mean distance between partners is 76 Km, which in a UK context represents approximately one hour’s driving time. This observation does not suggest a pattern of exclusively local links between SMEs and universities as the distance between partners varied considerably from firm to firm. For example, Indigo Consultants’ collaboration with Brunel University involved the largest distance, with the partners separated by 340 Km. In contrast, Motor-Tech and Magenta-Design were located only around 1 km from their university partners.

Despite the apparent existence of spatial proximity between the actors, this was not highlighted as an important factor in the formation or function of these links in contrast to the highlighted importance of spatial proximity in the extant literature (D’Este, Guy, and Iammarino 2013; Johnston and Huggins 2016a, 2016b). Instead, in the formation phase, the physical distance between the partners was often determined by the existing spatial scope of the SMEs’ networking activities, which typically focused on actors within their immediate geographic area. Therefore, a network focused on local actors resulted in the SME partnering with a local university and vice versa. The spatial profile of the collaborative links, therefore, did not result from a deliberate strategy to choose partners who were geographically close but was the result of the nature of the networking activities of the SMEs. This is described in the two quotes below:

I suppose that’s why [they’re all] local really, ‘cause I do a bit more networking locally [Magnolia Support].

So we had a relationship through a couple of people at the University, based on our marketing background locality, Bristol is probably forty minutes away from us [YellowSoft].

Where larger distances between actors was observed, the collaboration typically resulted from a link with an actor that was well established and long term in nature. In terms of the function of the projects, spatial proximity did not emerge as an important factor. Instead, as illustrated by the following quotes, interaction between the SME and their university partner was dependent upon the needs of the project rather than the physical distance between the partners.

If we want to call a meeting, people are prepared to travel. We’ve been over there a few times, they come over here, so it’s not … you know, it’s obviously easier if you were a bit closer but it’s a do-able kind of distance [Arc-Tech].

It was [the KTP advisor] that basically said, oh no you want to keep it local. Yeah, that’s wrong. The reason … but I understand the reasons why it’s right because you’ve just explained them. There are reasons of proximity that have lots of advantage. But that’s not the approach. The approach is where the best university in the country that has the best academics in the country that could realise this technology [Black Electronics].

Therefore, despite arguments within the extant literature that spatial proximity allows of the observation and assessment of their effectiveness (Wood and Parr 2005; Gulati 2007; Capaldo and Petruzzelli 2014) increasing the intensity of collaborative links (Storper and Venables 2004) the continued focus on the spatial proximity as a determinant of SME-university collaborative links may be misplaced. Hence it is proposed that:

Proposition 1: The spatial proximity of partners is not a direct determinant of the formation or function of SME-university links but is the outcome of existing spatial scope of SME networks.

4.2. Social proximity

The analysis suggests that social proximity plays a key role in the formation of SME-university collaborative links since SMEs’ partner selection process is underpinned by their existing network. Therefore, as is clear from the following quotes, SMEs focus on collaborating with university actors who are existing contacts or connected to them through existing contacts.

[The academic was] an old university colleague, so it’s a network, and I had … I lost touch for a while, but he turned up and we’re back in touch and the idea formed from that [Patient-Tech].

I’d just known [the academic supervisor] for a long time there. I just knew he knew a lot about play. He’d written a lot of papers. He’d speak at conferences … I’d subsequently been introduced some people, because they all engaged with us there, who were all of interest [Magenta Design].

[I have] lots of friends as professors and such like, you know, so, you know, at a social level [Blue Finance].

[The academic partner had been] involved in some of the initial projects that we’d worked on … and we’d sort of developed a relationship and … it just evolved from there really [Mine Tech].

While direct links with universities are typically based on previous interactions, the above quotes highlight that they are not necessarily formal ties. For example, direct links may also form through study links (contacting ex-lecturers), collaborating with friends in academia, exploiting links with academics developed through membership of trade associations, and attendance at conferences or industry events. Furthermore, existing contacts could represent long standing relationships. Where SMEs relies on indirect links, the search process utilises actors within their existing network to search for a partner to generate potential leads or using them as a sounding board to uncover potential partners, highlighting the importance of both strong and weak ties, as illustrated by the following quotes:

One of our scientists did his post-doc [at the partner university] and thought that there was somebody there who he was a student of who would be particularly good at advising us as to how to go forward. So they had a personal contact there [Gen Tech].

My wife was working at that University in a different department and we’d heard about the KTPs, so it seemed a … good way to get somebody to actually develop this software [Motor Tech].

[A partner from a previous project] kind of mentioned this KTP scheme and introduced us to the KTP representative [Red Soft].

Social proximity therefore facilitates the formation of SME-university links through enabling access to university actors. Therefore, it is typically through utilising existing network resources that SMEs identify a suitable university partner. This approach to collaborative tie creation consequently resembles the process of effectuation, relying on utilising the available means to develop ties (Sarasvathy 2001). Furthermore, social proximity facilitates the development of collaborative links through both direct links, based on strong ties where the actors are already acquainted, and indirect links, through weak ties where the actors have common connections. One respondent highlighted the social nature of the search process as follows:

So I mean we looked around at where the best universities were for Statistics and Maths … by asking around [my network]. Nothing scientific. No modelling involved, I’m afraid. Just by asking around where the good universities were [Purple Media].

Consequently, these findings give credence to the assertion that SME-university collaboration is inherently social in nature, as it is this aspect that promotes their formation (Eisenhardt and Schoonhoven 1996; Bjerregaard 2010; Steinmo 2015). Importantly, the social proximity observed here is more relational than cognitive in nature, as it enables contacts to be made between actors rather than promoting understanding (Steinmo and Rasmussen 2018). The importance of networking is outlined by the following quote, where one MD highlights a key aspect of his role is to continuously develop contacts:

I mean, my job’s to network … you don’t show, you don’t know [Blue-Design].

Therefore, these findings lead to a theoretical proposition that suggests that social proximity enables connections to be made between SMEs and universities:

Proposition 2: Social proximity is a key determinant of the formation of SME-university collaborations providing SMEs with a means of accessing a potential university partner.

4.3. Technological proximity

Technological proximity is important to both the formation and function of SME-university collaborative links as it facilitates the firms’ understanding of the university partner’s potential during partner selection and the absorption and use of the partners’ knowledge during the project. Therefore, the evidence confirms prior findings that technological proximity is important in facilitating the SMEs’ understanding of university actors in the formation and function of collaborative links (Marrocu, Paci, and Usai 2013; Chen and Xie 2018). The following quotes illustrate how technological proximity was manifest in a mutual understanding of the problem and a clear understanding between the parties:

You felt comfortable with [the academic supervisor] because you knew he knew what had to be done [Marine Consultants].

So [the academic supervisor] worked as a lecturer and he worked as a lecturer on business programmes, but his own background was as a consultant. He came from our world [Edu Tech].

We had a good relationship with regards to [communication and understanding]. It was always very understandable. We had no issues around that [YellowSoft].

The findings suggest that during the formation of collaborative links, SMEs with higher levels of technological proximity to their university partner were able to perform a more thorough assessment of their potential partners’ likely knowledge, expertise, and, accordingly, contribution to the project. Consequently, greater commonalities between the two parties’ knowledge base results in a more detailed discussion between the actors of the aims and objectives of the project, how it may be undertaken, and allows a more detailed specification on exactly how the university partner could contribute to achieving these. As a result, lower levels of technological understanding are associated with a more superficial assessment of a potential university partner, examining their affiliation, title, or qualifications, factors which simply indicated a basic outline of the academic partners’ knowledge and expertise as these SMEs lack a sufficient understanding to undertake a more in-depth assessment. The following quotes provide examples of how the SMEs perceived their potential partners and how they were able to understand their potential contribution due to technological proximity in terms of a shared knowledge base:

[The academic partners] did make sense. We’re a fairly technical company, and the specific esoteric around data analysis, data science, I think they were quite good at explaining in terms that we could understand. [Pink Soft].

There was no problem [understanding the academic partner]. They were, as a group, pretty business focused. I mean, they were very, very data driven and we were only a bit data driven … but we knew what each party wanted out of the whole thing. Fundamentally … we were pretty well aligned [Magnolia Support].

[The academic supervisor was] an individual that is extremely direct and very clear, and so I never felt academic jargon was an issue for him [Green Soft].

Importantly, technological proximity may also evolve during a project as while differing levels of understanding of the university partner were evident during the formation stage, during the functioning of the projects the SMEs reported few issues in terms of being able to understand and make sense of their academic partners’ words and actions. Therefore, during the operation of the project, overlaps in the actors’ knowledge and expertise were sufficient to allow a mutual understanding to develop during the projects, regardless of whether the SME possessed this at the outset. The evolution of technological proximity in the course of the project is illustrated by the following quotes:

If they sent something we didn’t understand, then I would ask and they would be quite good at making it clear what they were talking about. So, yeah, [Understanding the academic partner] was never really a problem. The whole point was obviously they’ve got a knowledge base that you haven’t got [Magenta Design].

[Sometimes] it seemed like there might be two or three languages through which [the knowledge] needed to be translated, rather than just one … it’s also where [the academic’s] input as the academic supervisor [was needed most] [Indigo Consultants].

The observation that technological proximity may evolve as the project progresses and actors within SMEs learn from those within the university corroborates prior assertions surrounding the dynamics of proximities (Balland, Boschma, and Frenken 2015; Lauvås and Steinmo 2019). Therefore, technological proximity is not necessarily a binary construct, either existing or not, but continuous in nature and existing on a spectrum. Furthermore, given the potential for technological proximity to evolve during the course of a project, a key implication is that it does not necessarily have to be complete at the outset of the project. Indeed, some SMEs recognised that their technological proximity with universities would never fully converge given that the academic partners were the experts in the field. However, where it is more developed at the start of a project, SMEs possesses a greater ability to perform a more in-depth assessment of the likely contribution of the university actor.

Within SMEs, university collaboration promotes learning which also promotes the development of technological proximity between the partner, underscoring why previous collaborations tend to facilitate repeated links as the actors learn to understand one another (Hewitt-Dundas, Gkypali, and Roper 2019). Importantly, the results suggest that technological proximity does not necessarily equate to a complete overlap with the university partners’ knowledge, but an understanding of it. For example, the following quotes from two of SMEs outline these issues:

It’s just that [the academic partner had] done it lots of times and they probably think to themselves well, it’s obvious ‘cause it’s called a knowledge transfer; but unless you understand what that means … somebody needs to be their guide through it or vice versa, and it has to have a roadmap in it. [Brown Media]

We found it was tricky conveying some of the complexities of our issue and our problems to them but, in fairness, I think, in hindsight, it was probably the right balance because they shouldn’t be bogged down by our problems. They had a mission to accomplish and sometimes we did feel as though they didn’t understand, but actually, in hindsight, had they spent a lot of time trying to understand. [Data-Tech]

Consequently, technological proximity only needs to be sufficient for SMEs to understand and absorb their partner’s knowledge; the existence of technological proximity does not necessarily mean both actors’ knowledge bases are homogenous. The arguments presented in this Section are captured in Proposition 3:

Proposition 3: Technological proximity has a positive influence on both the formation and function of SME-university collaborations and may also evolve as the collaboration progresses.

4.4. Organisational proximity

Organisational proximity is important to the formation but crucial to function of SME-university links. Most of the SMEs referred to previous experience of working with university actors, either formally or informally, at the outset of the project as an important determinant of working together successfully. At the outset of the project, it was this familiarity that led to the SMEs deciding to collaborate with a university. Importantly, the absence of organisational proximity does not preclude the formation of a collaborative link as asits presence appears to be a sufficient but not a necessary condition. Therefore, SMEs without prior experience of working with a university are still able to form a collaborative link.

However, organisational proximity was more important to the function of the projects as there was a realisation among the SMEs that to work effectively with university actors, they had to familiarise themselves with their working practices and project roles. Therefore, the development of organisational proximity during a project is not instantaneous, instead developing as the partners learned to work together. Additionally, where SMEs were already familiar with their partner or universities, there was still a need for the project teams to develop a working relationship specific to that project.

The development of organisational proximity involved first coalescing around the project and its objectives then organise themselves into a pattern of work. Consequently, organisational proximity may exist at the outset and subsequently evolve and grow as the collaboration progresses suggesting that it is continuous rather than binary in nature. This is process is described in the following quote:

You know, probably after three months we knew what everybody did and what the good bits were and the bad bits were and you work round that. So I think that was fine and that was good. And it was a strong team. You know, it was … everybody really knew where it was going, but maybe had slightly different opinions how it was going to get there [White Electronics].

Furthermore, the evolution of organisational proximity was not simply a case of SMEs adopting the working practices of their partner but both actors adopting a similar approach to the project. As a result, SMEs typically identified university actors who were ‘business focussed’, concentrating on the commercial aspect of the project rather than just the science underpinning it, as the drivers of successful collaborations. As such, organisational proximity cannot be considered as only related to working practices but a combination of shared motivations, practices, and timeframes. The broad nature of organisational proximity is highlighted in the following quotes:

There’s a difference between our researcher and the University’s researcher … So there are learning curves, I think. There’s not a [method] for establishing guidelines and principles of protocols for practice where, you know I think there could be a kind of naivety where you have to recognise that there are different interests at work, but then negotiate the sort of middle ground that makes it kind of effective for both parties, understanding what both parties’ interests are [Arc-Tech].

The only problem, I think, was just the sort of timeframes to manage projects and outputs really [Cyan Soft].

[The academic partner was] obviously very focused on what they perceive the nature of their discipline or their study, which is good, you know … Our perception is broader, probably because we’re having to stay in business, so having to succeed commercially as well as having kind of aspirations in terms of the quality and the vision of our business. But I think if you can … again, it’s a process of having a pragmatic and a realistic dialogue about that and being very open about that and being able to then acknowledge that [Red-Soft].

Organisational proximity can complement technological proximity as its existence facilitates understanding between actors. Therefore, for SMEs, effective knowledge transfer is influenced by both the similarities in the knowledge bases and similarities of the methods of working and motivations with their partners. Therefore, the findings related to organisational proximity not only corroborates existing evidence (D’Este, Guy, and Iammarino 2013; Johnston and Huggins 2016b; Kuttim 2016; Johnston 2020), but also reveal a more nuanced influence. First, organisational proximity developed through prior experience of working with universities allows SMEs to understand the potential contribution they can make to the project. This familiarity leads SMEs to consider university collaboration as a potential solution to the innovation problems they faced. However, as SMEs are also able to develop these links in the absence of organisational proximity then it may not be crucial for the formation of all linkages.

Given the fact that SMEs required an effective way of working with university actors in the function of their projects, the results show that organisational proximity can develop as the actors learn to work effectively with each other. Therefore, this also displays evolutionary tendencies, as the act of collaboration teaches SMEs how to do it. In addition, the results also highlight how organisational proximity may complement technological proximity in assisting with both understanding and implementation of knowledge. These assertions are formalised in Proposition 4:

Proposition 4: Organisational proximity has a positive influence on both the formation and function of SME-university collaborations and may also evolve during a project.

5. Conclusions

The findings presented in this paper add to the understanding OIS through SME-university collaboration by utilising the proximity matrix to assess the influence of cognitive and relational factors as the boundary conditions to facilitate these linkages. Through presenting a holistic examination of the SME-university collaboration process, the findings demonstrate that, overall, proximities are an important part of the formation and function of these collaborations. Therefore, the paper furthers the understanding of the OIS process by highlighting the role of proximities as important factors for assessing and understanding the boundary conditions that facilitate SME-university collaborations.

Importantly, these results show that the boundary conditions that promote the formation and function of SME-university collaborations are created by a combination of proximities rather than a single aspect. Furthermore, as the results do not suggest the primacy of a single proximity, it appears pertinent to take a broad view of these boundary conditions, thereby justifying the proximity matrix approach.

The findings not only highlight the importance of proximities but also reveal complementarities between the proximities; for example, technological proximity complements social proximity. This complementary effect means that having identified a university actor as a potential partner, possessing a similar knowledge base allows SMEs to comprehend the extent of the potential partners’ knowledge and expertise and how it relates to their project.

In line with the OIS approach that examines the antecedents of OIS practices in terms of the individual, team, and organisational levels, the findings also highlight which of these are influenced by each proximity. For example, social proximity is observed at the individual level through connections between people, whereas organisational and technological proximities are discussed in terms of individuals, groups and the organisations’ ability to understand one another and work together effectively. Therefore, when discussing the boundary conditions and OIS practices it is important to note that these can be understood in terms of a range of proximities whose influence is observed at either the individual, team, or organisational level. Indeed, these findings go some way towards answering calls to further the micro-level antecedents and contingencies that underpin OIS practices (Beck et al. 2020).

Therefore, this paper contributes new insights into the phenomenon of SMEs engaging in OIS through university links by highlighting that SMEs’ ability to engage in university collaborations hinges on their ability both access and understand university actors through a range of cognitive and relational factors. SME access to university partners is facilitated through social proximity, and their ability to understand their partner through technological and organisational proximity. Importantly, these findings also highlight the nuances of these proximities in SME-university collaborations, outlining their distinct contribution to the specific stages of the collaboration. For example, social proximity contributes to the SMEs’ successful search for a university partner, while technological and organisational proximity allows the SMEs to both assess the usefulness of a university partner and enables the project team to function smoothly. Therefore, these findings yield important insights that allow the development of four theoretical propositions, summarised in Table 2.

Table 2. Proximity matrix – Summary of propositions on the role(s) of each proximity

Proximity	Theorised Role in SME-University Collaboration
Spatial Proximity	The spatial proximity of partners is not a direct determinant of the formation or function of SME-university links but is the outcome of existing spatial scope of networks.
Social Proximity	Social proximity is a key determinant of the formation of SME-university collaborations providing SMEs with a means of accessing a potential university partner.
Technological Proximity	Technological proximity has a positive influence on both the formation and function of SME-university collaborations and may evolve as the collaboration progresses.
Organisational Proximity	Organisational proximity has a positive influence on both the formation and function of SME-university collaborations and may also evolve during a project.

The paper also presents further theoretical insights into the nature of proximities and university-industry collaboration. Firstly, the results have highlighted the continuous rather than binary nature of proximities. Therefore, it is not necessarily a case of whether a proximity is present or not, but a case of the degree to which it is present. Secondly, the completeness of these proximities is not necessarily a limiting factor as they may evolve over the course of a project therefore, their absence is not a limiting factor as collaborations can occur in their absence and they may develop as the project progresses. Therefore, the incomplete or limited proximities between actors do not prevent the formation of SME-university ties.

Given these findings, the process of SME-university collaboration is clearly a learning process for the firms involved, offering explanation to previous findings that once a firm begins collaborating with a university it is likely to once more (Hewitt-Dundas, Gkypali, and Roper 2019). Indeed, as the broader literature suggests that processes of open innovation and inter-organisational collaboration are increasingly predicated on the cognitive and relational factors (Huber 2012; Steinmo 2015; Steinmo and Rasmussen 2018), it is important to examine these in terms of OIS and boundary conditions.

The theoretical propositions outlined in this paper here therefore provide a testable framework for future research examining OIS through SME-university interaction. As this paper looks solely at UK SMEs in a broad range of sectors, the framework could be utilised to examine specific geographies or sectors so that generalisable patterns surrounding the importance of proximities to the formation and function of SME-university collaborations may be established. Furthermore, larger scale datasets could be used to establish causal relationships between the existence of these proximities and the outcomes of SME-university collaborations. Finally, while the operationalisation of the proximities within the paper identifies their roles and the stages of the project at which they are important, additional work is required to quantify the similarities between actors within SMEs and universities. This could then also yield insights into optimal levels of proximity and the facilitation of collaborations and their successful outcomes as well as the linear or non-linear nature of the relationship. As such, this may yield insights into when the closeness of actors may become dysfunctional.

While this paper has highlighted insights from university collaboration on SMEs, the OIS framework outlines the fact that outcomes can be created along the entire spectrum of the scientific process (Beck et al. 2020). While activities such as joint patenting and publications have been examined (Wong and Singh 2013; Murgia 2018), future research may wish to focus on the impacts of SMEs on universities through the examining the extent to which these collaborations may facilitate the development of datasets, publications, and teaching materials.

Beyond theoretical implications, the findings presented here are also relevant to the policymaking and SME communities. With respect to policymakers, encouraging SME-university links should not consist of merely encouraging interaction but encouraging the development of both cognitive and relational similarities. Therefore, policy initiatives should first promote access to universities on the part of SMEs through encouraging interaction through networking. In addition, policies should also advocate communication between these actors to promote an understanding of university knowledge among SMEs. In doing so, policymakers must appreciate the spatial scope of individual SME’s networks, promoting collaboration that fits with this as the spatial scope likely reflects the appropriate distance over which they will collaborate rather than political boundaries that delimit the policymakers spatial focus.

In terms of practice, these findings suggest that SMEs seeking to work with universities should utilise their present network resources to seek partners. In addition, familiarity through informal interactions may allow them to gain experience of a university’s knowledge base and methods of working to promote the development of technological and organisational proximity. Importantly, as these may evolve in the course of the project, SMEs need not seek a partner with homogenous knowledge or working practices but be able to understand these sufficiently to learn from them.

Acknowledgments

The author would like to thank the editors and reviewers of this Special Issue for their constructive comments in the review process. The paper has been much improved because of their input. As ever, all errors remain my own.

Disclosure statement

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