Multiplicity of alliance learning in the entrepreneurial process: strategies of early-stage biotech firms

Abstract

Entrepreneurial firms depend on knowledge variety and the ability to manage alliance-network learning for knowledge acquisition, which are both challenging. Some studies argue reliance on one key alliance partner is more effective for entrepreneurial firms with limited resources as it is less demanding than collaborations with multiple ones, while others demonstrate that alliances with different organizations significantly benefit them. Firm strategies and mechanisms of the alliance-network learning with multiple partners remain unclear, and illuminating this puzzle is relevant for understanding small and young firms creating innovations. Focusing on the early stages of human health biotech firms in Canada, this paper examines how they use the alliance-network to identify learning opportunities and pursue knowledge accumulation over their successive developmental stages. Adopting the multiple case study method and analyzing the focal firms’ collaborations and learning outcomes, this research advances a processual model of multiplicity of learning. The model identifies firm-specific strategies combining a variety of knowledge building mechanisms and multiple partners guided by organizational goals to innovate. The study contributes to the intersection of entrepreneurial development and alliance learning literatures with a novel view of early-stage firm learning strategies, and offers insights to entrepreneurs, small firms, and policymakers for innovation.

Résumé

Les firmes entrepreneuriales dépendent de la diversité des connaissances et de la capacité à gérer l’apprentissage des alliances et des réseaux pour l’acquisition des connaissances, qui sont toutes les deux difficiles. Certaines études avancent que le recours à un partenaire d’alliance clé est plus efficace pour les firmes entrepreneuriales aux ressources limitées, étant donné qu’il est moins exigeant que les collaborations avec plusieurs partenaires, tandis que les alliances avec différentes organisations leur apportent des avantages significatifs. Les stratégies des firmes et les mécanismes d’apprentissage des alliances et des réseaux avec des partenaires multiples ne sont toujours pas clairs, et il est pertinent d’éclaircir cette énigme pour comprendre les petites et les jeunes entreprises qui innovent. En se concentrant sur les premières étapes des entreprises de biotechnologie en santé humaine au Canada, cet article examine comment celles-ci utilisent le réseau d’alliances pour identifier les opportunités d’apprentissage et poursuivre l’accumulation de connaissances au cours de leurs étapes successives de développement. En adoptant la méthode des études de cas multiples et en analysant les collaborations et les résultats d’apprentissage des entreprises focales, cette recherche avance un modèle processuel de la multiplicité de l’apprentissage. Le modèle identifie des stratégies spécifiques aux entreprises combinant une diversité de mécanismes de développement des connaissances et de multiples partenaires guidés par des objectifs organisationnels d’innovation. L’étude contribue au croisement des littératures sur le développement entrepreneurial et l’apprentissage des alliances avec une vision nouvelle des stratégies d’apprentissage des entreprises en phase de démarrage, et offre des perspectives d’innovation aux entrepreneurs, aux petites entreprises et aux décideurs politiques.

Keywords:

• Entrepreneurial firm
• alliance network
• multiplicity of learning
• strategy

MOTS CLÉS:

• Firmes entrepreneuriales
• réseau d’alliances
• multiplicité d’apprentissage
• stratégie

1. Introduction

As the world confronts major health, environmental, and social challenges, many small and young companies are creating new solutions. Among others, biotech companies, many of them small startups, are developing vaccines and diagnostics products, often working in collaborations (Lee 2020). Entrepreneurial firms leverage successful alliance projects to build internal knowledge for innovation and growth (Kamuriwo and Baden-Fuller 2016; Macpherson and Holt 2007; Mesquita, Anand, and Brush 2008; Simba 2015). Collaborative projects bring valuable resources, expertise, and opportunities for learning (Dutta and Hora 2017; Gulati 2007; Powell, Koput, and Smith-Doerr 1996). However, learning from an alliance-network requires a firm’s resources and capabilities. There are risks and challenges in forming alliances properly, managing partnerships to achieve synergies, and seeking effectiveness in knowledge acquisition and adaptation (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018; Love, Roper, and Vahter 2014; Lowik et al. 2012; Moghaddam, Bosse, and Provance 2016).

In a context of limited resources, recent work discussed the collaborative ‘multiplexity’, defined as seeking several knowledge types (technological, market, and managerial) through one key partner to maximize the learning effects (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018). With only one partner, the ‘multiplex’ alliance can mitigate the alliance formation costs, help to navigate the partnership management easily, and better nurture the inter-partner relationship to achieve more learning. Studies on multiplexity validate the discussion on the strategies of entrepreneurial firms pursuing learning for innovation under constrained early-stage conditions. However, multiplexity does not address the learning scenarios when such firms engage multiple partners (Simba 2015).

The existing literature shows that early-stage entrepreneurial firms regularly engage in alliances with different partners to acquire various types of knowledge (Baum, Calabrese, and Silverman 2000; Kamuriwo and Baden-Fuller 2016; Pisano 2006). In the general activity pattern of the entrepreneurial process, ‘early stages’ refers to the opportunity identification and development period of the newly established enterprise dedicated to innovative product or service development before it achieves the first sale and operational viability (Chell 2013; Muñiz Ávila, Silveyra León, and Segarra Pérez 2019; Murray 2004; Pisano 2006). For science-based start-ups like dedicated biotech firms (DBF), it involves an iterative entrepreneurial process working with multiple partners in a prolonged period. For example, a DBF often needs to collaborate with university laboratories, hospitals, venture capital firms, contractual research organizations, large pharmaceutical companies, other biotech firms, and different levels of government during the 10–15 year first product development period (BIOTECanada 2020). Explorative activities and uncertainty characterize early stages, while the strategic direction may still be in the process of development and changing (Muñiz Ávila, Silveyra León, and Segarra Pérez 2019; Sarasvathy 2001).

Learning while navigating multiple partnerships creates crucial challenges to early-stage firm strategy, which remains largely under-explored (Martínez-Noya and Narula 2018; Simba 2015). For early-stage firms operating with constrained resources, inadequate experience, and incipient networks, partnering choices are often limited, and learning can be interrupted due to the project failure (Pisano 2006). Managing multiple partnerships can stretch already limited resources (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018). Learning for product innovation intertwines with learning for firm development, which complicates the process. The strategy to address these challenges can fundamentally affect learning as well as firm development (Macpherson and Holt 2007). To better understand the multi-partner scenario, this study focuses on the research question of how early-stage entrepreneurial firms strategize to manage learning with an evolving alliance-network.

This investigation analyzes cases of early-stage human health biotech startups learning with numerous collaborations. It offers evidence to support a new multiplicity of learning strategy of early-stage entrepreneurial firms in building important organizational knowledge for innovation while expanding their network of partners. This multiplicity strategy integrates four learning mechanisms, starting with ‘accessing’ and ‘transferring’ knowledge from initial partners, and then evolving to ‘catalyzing’ and ‘disseminating’ emerging knowledge from new partners. Adopting the multiplicity strategy allows early-stage firms to pursue various alliances effectively for richer learning opportunities that fit their needs with evolving organizational and inter-organizational resources and capabilities. This study advances the alliance-network learning literature, going beyond the multiplexity view that involves only one main partner for small firms toward an alternate multi-partner learning strategy.

2. Alliance-network learning of entrepreneurial firms

Learning in entrepreneurial firms occurs in a context of constrained resources, incipient assets, low survival rates, and major strategic hurdles (Ambos and Birkinshaw 2010; Burton and Beckman 2007; Graebner 2004; Simba 2013; Van de Ven and Poole 1995). These constraints inhibit learning in early-stage entrepreneurial firms. Low recognition and weak networks limit nascent firms’ choices of partners in business and innovation activities deemed essential for success (Martínez-Noya and Narula 2018; Powell, Koput, and Smith-Doerr 1996). In addition, the partner often does not support the learning, instead, exploiting the intellectual property of the new entrepreneurial firm (Alcacer and Oxley 2014; Inkpen and Tsang 2007; Kale, Singh, and Perlmutter 2000; Moghaddam, Bosse, and Provance 2016). Furthermore, early-stage entrepreneurial firms’ high-risk exploratory innovation could even lead to the firm’s demise (Muñiz Ávila, Silveyra León, and Segarra Pérez 2019; Pisano 2006). Entrepreneurial firm knowledge acquisition and learning in early stages can encounter challenges due to liability of newness and small size, requiring intentional efforts and managerial actions to succeed (Macpherson and Holt 2007; Simba and Ndlovu 2014).

To drive product innovation, firms must build organizational knowledge, balancing exploration and exploitation in learning (Hargadon and Fanelli 2002; March 1991). Firms bear, deploy, combine and re-combine existing knowledge to develop new products (Amin and Cohendet 2004; Chandler 1992; George 2005; Grant 1996; Kogut and Zander 1992; Leonard-Barton 1992). Through collaboration with other organizations, knowledge embedded in various locations of a network is accessed and used by the focal firm (Figueiredo, Larsen, and Hansen 2020; Grant and Baden-Fuller 2004; Lipparini, Lorenzoni, and Ferriani 2014; Rothaermel 2001; Saxenian 1994; Simba 2013). Using knowledge from the network is a source of competitive advantage (Gulati 2007; Lavie 2006; Lipparini, Lorenzoni, and Ferriani 2014; Smith, Collins, and Clark 2005). Networks of organizations are an essential feature of innovation dynamics, affecting new knowledge accumulation at the firm level (Ingram 2002; Powell, Koput, and Smith-Doerr 1996; Saxenian 1994; Simba 2015; Zollo, Reuer, and Singh 2002). In reciprocal alliances, firms access the partners’ expertise, share experiences, and exploit complementarities to generate new knowledge accruing to both parties (Grant and Baden-Fuller 2004; Lubatkin, Florin, and Lane 2001; Macpherson and Holt 2007; Mowery, Oxley, and Silverman 2002; Simba 2013). The gained knowledge contributes to both new product development and managerial efficiency such as relationship management.

Entrepreneurial firms often engage a diverse alliance-network in their early-stages (Kamuriwo and Baden-Fuller 2016). The literature emphasizes the importance of learning from a variety of partners in multiple domains (or types) of knowledge, including technological, market, and management (Macpherson and Holt 2007; Penrose 2009; Rothaermel 2001; Simba and Ndlovu 2014). Diversity and variety in learning are critical for both innovation and other organizational objectives (Garg and Zhao 2018; Gronum, Verreynne, and Kastelle 2012; Leiponen and Helfat 2010). Greater breadth in the type of knowledge sources, such as suppliers and customers, contributes to more successful innovation at the firm level. Multiple sources of knowledge are better to address the cognitive biases in a firm’s search process, contributing to the cumulative learning, that is, depth of search (Leiponen and Helfat 2010). This enables the firm to generate more innovation outputs from external collaboration (Love, Roper, and Vahter 2014). Different partnerships can facilitate various applications of knowledge in innovation (Powell and Grodal 2006; Simba 2015). Furthermore, applying one type of knowledge often requires the combination of complementary types of knowledge to ensure comprehensiveness and applicability (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018). Relationships with various partners can enable multiple types of knowledge in pursuing innovation and growth in entrepreneurial firms.

While recognizing the importance of partner variety, the literature also notes challenges that constrain entrepreneurial firm learning in complex collaborations applicability (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018). One challenge is the limited absorptive capacity due to the lack of prior knowledge constrains entrepreneurial firms from learning effectively (Cohen and Levinthal 1990; Yoo, Sawyerr, and Tan 2016). Another challenge is that short inter-organizational relational capabilities can limit learning from partners (Kale, Singh, and Perlmutter 2000; Lorenzoni and Lipparini 1999; Lowik et al. 2012; Zollo, Reuer, and Singh 2002). Developing such skills and capabilities requires time, while firms make internal R&D investments, coordinate and organize tasks, and build relationship trust. Despite the advantage of learning from a variety of partners in the early stages, many are unable to benefit due to the inherent challenges and the possibility of failure (Pisano 2006).

Research on the inherent challenges in alliance learning emphasizes a strategy based on ‘multiplexity’, that is, acquiring different types of knowledge with one or a small number of key partners (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018; Moghaddam, Bosse, and Provance 2016; Yoo, Sawyerr, and Tan 2016). It argues for transferring multiplex knowledge via the strengthened relationship with a key commercial partner (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018; Lowik et al. 2012). According to this view, this strategy allows entrepreneurial firms to reconcile the tradeoffs between the value of a partner and knowledge variety while addressing learning management complexity (Moghaddam, Bosse, and Provance 2016). Research on learning multiplexity stresses the managerial complexity of a diverse alliance network over its potential benefits and suggests effective learning from a single partner, rather than various partners given the limit of relational capabilities (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018).

While the multiplexity view captures the complex challenges that alliances involve and one strategy to address them, it raises three issues to be explored further. By focusing exclusively on knowledge variety from one partner, it does not examine alliance learning with multiple partners. Entrepreneurial firms, however, engage in a diverse alliance network with varied partners dealing with multiple challenges simultaneously for competition (Simba and Ndlovu 2014). The gains of learning efficiency from having only one partner may sacrifice the benefits from new knowledge variety creation, a more important goal for innovation-focused firms. The multiplexity strategy indicates an emphasis on efficiency and exploitation, giving less attention to exploration in learning (Lavie, Stettner, and Tushman 2010). A single-partner knowledge variety strategy, or ‘multiplexity’, may not mitigate the thwarting of knowledge creation due to limited search breadth and inevitable biases associated with lacking partner diversity (Leiponen and Helfat 2010; Martínez-Noya and Narula 2018).

In addition, the multiplexity approach views the knowledge the focal firm acquires from a collaboration as given by the partner, ready for transfer and adaptation. Complex learning, however, beyond transferring and adaptation occurs in alliances (Inkpen and Tsang 2007; Martínez-Noya and Narula 2018). Developing and integrating knowledge characterizes the entrepreneurial organization’s objectives (Simba 2015). There is a need to consider the conceptual shift of knowledge from a ‘transferable and storable asset’ to a ‘process of searching for and recognizing’ what and how to use it in a specific firm’s context (Macpherson and Holt 2007, 185). Therefore, research is needed to understand better whether multiplexity captures the processes in innovation-driven contexts, where the pressure for novel knowledge creation is central to developing new products, such as in an early-stage entrepreneurial firm.

Moreover, there is still limited understanding of how entrepreneurial firms pursue learning in collaborations considering the complex challenges, and specific organizational objectives and responses to this complexity (Martínez-Noya and Narula 2018). Existing studies show small entrepreneurial firms often conduct alliance learning in their early stages with different firms and organizations to create knowledge and overcome resource constraints, rather than relying on variety from one key partner (Powell and Grodal 2006; Simba 2013; Simba 2015). Recent research of interactive learning strategies for innovation in the biotech sector focused on large multinationals (Figueiredo, Larsen, and Hansen 2020). While it is known that collaborations significantly affect innovative and organizational capabilities, there is a need for more studies on how young entrepreneurial firms manage the complexity of multiple partners over time or across stages. Given this gap in understanding learning beyond single partners, this study examines alliance learning situated and unfolding in a context that includes the focal early-stage firm, its partners, and their collaborative relationship (Ingram 2002; Inkpen and Tsang 2007; Macpherson and Holt 2007). It extends the conceptualization of alliance learning under constrained resources toward integrating the entrepreneurial development literature into an innovation-driven process model. This contribution increases understanding of how the entrepreneurial firm’s capabilities and context evolve over time as it develops, partnerships evolve, and new collaborators emerge.

3. Methods and data

3.1. Research design, setting, and sample

This study uses in-depth multiple case studies to investigate how entrepreneurial firms build knowledge from collaborations with various partners (Davis and Eisenhardt 2011; Eisenhardt 1989a; Gulati, Lavie, and Singh 2009; Langley 1999; Yin 2003). A qualitative approach provided a richer understanding of early-stage firms learning strategies with their multiple partnerships. Gaining access to multiple key informants per firm to produce richer data for each firm’s partnerships was essential as detailed information was not easily available. The study of process involves historical and evolutionary data as the young firm undergoes different stages made possible with a relatively smaller sample (Crick 2021). The field data support the comparative analysis from rich information on 26 collaborations of 7 DBF located in Montreal in the Quebec province, Canada (Table 1) (Eisenhardt 1989b; Eisenhardt and Graebner 2007). The analysis includes all the collaborations for each company with firms and other types of organizations that generated new knowledge and multiple types of learning for each partnership.

Table 1. Sample of early-stage biotech companies and collaborations.

Company	Number of collaboration sampled	Number of employees	Therapeutic area	Initial R&D stage	R&D stage at interview	Number of products
						Preclinical	Clinical
Alpha	6	48	Oncology & diabetic neuropathic pain	Target discovery	Clinical phase 2	1	3
Bravo	3	67	Endocrine, nutritional & metabolic diseases	Lead candidate	Clinical phase 3 and marketing	1	1
Charlie	3	60	Hepatitis C; HIV	Lead candidate	Clinical phase 1	3+	4
Delta	2	55	Oncology	Target discovery	Clinical phase 2	2	2
Echo	3	37	Proteomics service	Target discovery	Third-generation of specialized target discovery	–	–
Foxtrot	5	7	Functional dairy products	Target discovery	Clinical phase 2	7	2
Golf	4	11	Ovarian and breast cancer; bone loss	Target discovery	Target platform and preclinical development	3	0

Notes. 1. We use the code words system of the International Civil Aviation Organization (ICAO) to represent our participant firms.

2. Sample of collaborations (CoL): N = 26 (CoL#1∼#26). Each collaboration involves more than one learning mechanism, so that our total count of learning mechanism samples is 40. The 26 collaborations are numbered according to the order of the firms they belong to, from Alpha to Golf. For example, CoL#1 ∼ 6 belong to Alpha, CoL#7 ∼ 9 belong to Bravo, CoL#10 ∼ 12 belong to Charlie, CoL#13 ∼ 14 belong to Delta, CoL#15 ∼ 17 belong to Echo, CoL#18 ∼ 22 belong to Foxtrot, and CoL#23 ∼ 26 belong to Golf.

3. In sampling the collaborations, we count multiple collaborators of one kind as one collaboration. For example, all sampled firms involve several to a large number of CROs, university collaborators, or KOLs, while they are counted as one in this table.

4. All companies in the sample were private except for Firm Bravo, which had already gone public a couple of years prior to our study.

Early-stage firms in the biotechnology sector present an ideal research setting for our study. The literature establishes the high presence of entrepreneurial startups and inter-organizational collaborative networks among biotechnology firms (e.g., Baum, Calabrese, and Silverman 2000; Inkpen and Tsang 2007; Owen-Smith and Powell 2004; Powell, Koput, and Smith-Doerr 1996; Simba 2013). Like other technology ventures, biotech startups often begin with limited resources. Firm-level innovative capabilities beyond basic research, such as preclinical and clinical studies, and marketing expertise, are crucial to advance products, attract partners, and foster growth (Colombo, Meoli, and Vismara 2019). The lengthy, expensive, and risky product development process (typically 12-15 years and up to U$500 million per discovery) makes any small biotech firm reliant on external knowledge and capital of universities, research institutes, investors, and other bio-pharma companies (Dutta and Hora 2017; Pisano 2006; Robbins-Roth 2000). The ability to form, manage, integrate, and learn from partnership relations is essential (Baum and Silverman 2004; Colombo, Meoli, and Vismara 2019; Powell 1998; Simba and Ndlovu 2014).

The selection of DBF in Montreal, Canada is due to their significant presence among the largest hubs of North America, with many early-stage companies (Deloitte 2018; Gertler and Vinodrai 2009; Niosi and Bas 2003; Tremblay and Klein 2014). The Montreal context facilitates access to start-ups with knowledgeable managers to collect data and obtain sufficient depth of information. The Montreal biotech cluster has over 500 companies, employing 28,150 people (BioQuebec 2010). It is the home of many early-stage firms and established pharmaceutical companies thanks to the ongoing municipal policies in promoting innovation (Niosi and Bas 2003; Schiffauerova and Beaudry 2012; Tremblay and Klein 2014). National surveys find 70% of Canadian biotech firms having fewer than 50 employees in 2009, and 56% having fewer than 25 employees in 2018; collectively, small biotech firms account for higher R&D expenditures over larger companies in Canada, and the Montreal cluster remains the top in terms of number of investors, patents, and collaborations (Deloitte 2018; Gertler and Vinodrai 2009; Schiffauerova and Beaudry 2012). In 2018, 67% of Canadian biotech firms identified themselves in the discovery or emerging phase of development, and the majority (56%) are members of the health biotech subsector (Deloitte 2018). Based on the last robust national survey, Quebec-based human health firms, mostly in Montreal, formed about 3.3 alliances (formal collaborative agreements) per firm with different types of companies or research institutes (Statistics Canada 2005). This choice of context is empirically rich for studying learning and collaboration practices in early-stage firms with experienced informants (Crick 2021; Simba 2015).

Purposeful sampling guided the selection of firms and collaborations (Patton 2002). Focusing on the early-stage entrepreneurial firms’ learning processes for innovation, the sampling rules included: independent human health DBFs in new drug or novel platform development; did not have the first product approval for marketing and sales by the regulatory authorities, such as Food and Drug Administration (FDA) in the U.S.A. and Health Canada in Canada; and small-sized with less than 50 employees (Table 1). All the selected firms were actively involved in collaborative activities. Investigation included how they advanced their technologies and products along the key milestones in the product innovation process.

Using the above sampling rules, a list of two dozen Montreal-based human health-related firms were identified from the Canadian Life Sciences Database (BIOTECanada 2009). Despite the challenges of recruiting small, privately-held, high-tech firm managers for interviewing, fourteen of the two dozen early-stage firms accepted our invitation to participate. After the first interviews, seven firms accepted to engage in further in-depth interviews with three knowledgeable senior managers from founder to top management team. Interviewing senior managers is an effective method used in other studies on entrepreneurial firms (Simba 2015; Simba and Ndlovu 2014). The seven firm cases reached the theoretical saturation, according to the appropriate number of 4–10 cases for theory generation established in the literature (Eisenhardt 1989a , 545). As well, the data analysis adequately represented the emerging themes, in this case, the learning mechanisms and the strategy pattern (Saunders et al. 2018).

All inter-organizational collaborations at each of these early-stage companies where the informants stated that new knowledge emerged were included in the sample. Collaboration refers to all voluntary arrangements that involved durable sharing or co-development of new products and technologies between biotechnology firms, as well as with other upstream and downstream private and public firms, and research organizations (e.g. Baum, Calabrese, and Silverman 2000; Dutta and Hora 2017). The broad inclusion of partnerships covered equity and non-equity partners, and various types of other partners including funders, large pharmaceutical companies, other biotech companies, research institutes, universities, and contract-based partnerships. In the end, the total count of 26 collaborations for all seven firms were analyzed. All the 26 collaborations were exploratory in terms of new knowledge generation (Lavie and Rosenkopf 2006).

3.2. Data collection and analysis

The main data sources included interviews and archival materials. After eight pilot interviews for interview guide verification, twenty-one semi-structured interviews (90–120 minutes each) were conducted with senior managers. Managers interviewed included the firm’s founders, CEOs, chief financial officer (CFOs), chief scientific officers (CSO), as well as senior or middle-level scientific executives. All interviewees were the most knowledgeable about their firm’s history, strategy, and R&D activities. Multiple interviews with different informants at each firm insured diverse perspectives for richer data and triangulation.

To examine entrepreneurial firm learning strategies with multiple partners, the interview aimed for collecting data on each of the 26 collaborations regarding its characteristics, knowledge gained from each partner and process, and connection with other collaborations of the same focal firm. The interview protocol therefore included three sections of open-ended questions on both the firm- and collaboration-levels. The first section of questions focused on the firm’s history, resources, capabilities, and overall strategy. The second group of questions centered on the firm’s partnership portfolio, its partnership strategy, and a chronological partnership list. The third section asked informants to focus on each collaboration over time. Typically, the informants would freely discuss each partnership of the company according to the temporal order and considering the significance and learning processes of specific partnerships, focusing on the content, management, and knowledge consequences of each partnership.

Archival data included corporate materials and internal documents (e.g. business plans, company brochures, IPO prospectus) sourced from the participant firms, public databases, and the Internet (NEWSCAN.com and Sedar.com) as well as company websites. Public materials were very limited for the sampled firms as most were privately held. Archival documents, together with multiple interviews with different informants at each firm, contributed to data triangulation. This triangulation reduced the biases of retrospective data and informants (Patton 2002) and strengthened the accuracy confidence of findings (Jick 1979).

Data analysis identified new knowledge accumulation and learning, categorizing when the firm achieved a ‘change in knowledge base’, or obtained new knowledge, know-how, or ideas. The first step of data analysis involved write-ups for each firm’s development and collaboration history to increase familiarity with all the data, as well as to support within and cross-case analysis (Langley 1999; Simba 2015). Both authors cross-checked write-ups during and after the company revisit and data clarification. Tables and visual maps were produced to present the timelines and evolution for each firm. Subsequently, using NVivo qualitative data analysis software, data coding was conducted on the written cases interacting with themes from both the literature and the initial reading of the field data. The coding process identified four mutually exclusive learning mechanisms, used by the firm in a variety of sequences and combinations to achieve certain knowledge gains. While each of the 26 collaborations involved at least one learning mechanism, many encompassed multiple types of knowledge and multiple learning mechanisms, for a total count of 40 learning mechanism samples. This observation revealed a more complex learning process beyond what the literature indicates. Detailed accounts of how these learning observations interrelated were recorded too. Verification between codes and the data iterated, creating internal validity of the patterns identified.

4. Findings: multiplicity of learning in entrepreneurial firm partnership

The findings of this study indicate that the early-stage entrepreneurial firm manages its alliance-network relationships with various partners using a multiplicity of learning mechanisms. Pursuing its strategic goals of innovation, it can adopt the combination of four specific learning mechanisms during collaboration: knowledge accessing, knowledge transferring, knowledge catalyzing, and knowledge disseminating (Figure 1). As the focal firm uses, leverages, and explores the initial partner’s knowledge and network connections, it engages multiple partners in order to gain both diverse types of knowledge and richer knowledge of one type (Figure 2). The use of a specific mechanism emerges when new learning opportunities with the same, or more often, with new partners arises. An integrated model of this learning process is elaborated to show how the entrepreneurial firm with deliberate learning goals expands the partner interaction and network, and actively pursues more and diverse knowledge acquisition through various collaborators and multiple mechanisms. The following part first identifies and compares the four distinctive learning mechanisms, followed by the presentation of the multiplicity of learning in an early-stage firm.

Figure 1. Four learning mechanisms.

Source: Authors’ elaboration.

Figure 2. Multiplicity of learning: knowledge variety and partner variety.

Note. 1–5 represents the route of progress. Types of knowledge: technological, managerial, and market.

Source: Authors’ elaboration.

4.1. Four mechanisms of learning in alliances

An entrepreneurial firm engages in collaborative relationships that integrate four learning mechanisms (Figure 1(a–d)). They occur progressively during the focal firm’s overall learning process with the same or different partners. Knowledge accessing and knowledge transferring occur first to initiate the relationship and lay the foundation for knowledge catalyzing and knowledge disseminating. These four mechanisms are mutually exclusive in terms of the source, role of the partner, characteristics, and emergence of the knowledge type, which are summarized in Table 2.

Table 2. Comparison of four learning mechanisms.

Learning mechanism	Source of initial knowledge	Practice of the learning mechanism	Characteristics of new knowledge accrued to the focal firm
Knowledge accessing	Partner	Applying the sourced knowledge	Newly generated during the partnership; different from the sourced knowledge
Knowledge transferring	Partner	Internalizing the sourced knowledge	Equal or adapted from sourced knowledge
Knowledge catalyzing	Partner and partner’s network	Applying or internalizing the spillover knowledge that differs from the sourced one	Newly generated beyond the partnership; different from the sourced knowledge
Knowledge disseminating	Focal firm	Sharing and generating community responses and ideas on the focal firm knowledge	Newly generated during the partnership; complementary to the sourced knowledge

Note. IND: Investigational New Drug NDA: New Drug Approval; KOL: Key Opinion Leader CRO: Contract Research Organization.

Source: Authors’ elaboration.

The sampled focal firms use knowledge accessing and knowledge transferring, two well-studied mechanisms known for driving alliance formation (Grant and Baden-Fuller 2004; Kamuriwo and Baden-Fuller 2016). Consistent with the literature, knowledge accessing collaborations use the collaborators’ expertise (technological and/or market types) to create new knowledge. Alternately, by the knowledge transferring mechanism, the focal firms acquired and internalized the needed technologies (adapted if necessary) from the collaborator. Although accessing generates new knowledge for sharing with the partners (e.g. each party commercialized different parts of the generated patents), the transferring mechanism allows more internal knowledge accrued to the focal firm moving toward independent innovation (Figure 1(b)). Many informants stated that licensing, a typical knowledge transferring agreement, was ‘strategic’ or ‘useful’ in building their current capabilities. The focal firms used these two mechanisms in various product development stages, including laboratory research, various testing, new applications of the technology, and so on (Figures 1(a,b) and 3).

Figure 3. Evolving capabilities and main partner organizations of a typical dedicated biotech firm in the drug development process.

Source: Authors’ elaboration.

As the focal firm identifies gaps and opportunities for new learning, it uses knowledge catalyzing and knowledge disseminating mechanisms after initially accessing or transferring the partners’ know-how. Different from the planned use of the first two mechanisms, these two learning mechanisms are often emergent when the focal firm identifies new, unanticipated knowledge (aka. spillovers according to Owen-Smith and Powell 2004) due to a closer relationship and enhanced interactions with the partner. Knowledge catalyzing refers to as the focal firm leveraging the spillover knowledge that emerges during collaborations. Facilitated by the partner, catalyzing goes beyond what is expected in the formal partnership agreement or the current partnership (Figure 1(c)). Using the initial formal agreement as a catalyst for more interactions and new partners, focal firms identify and leverage a broader range of scientific and market know-how not accessible initially. Catalyzing advances the focal firm’s knowledge to a new level in interaction with the partner or the partner’s network.

Differently, knowledge disseminating refers to the mechanism observed when the focal firm shares its proprietary knowledge or materials with the scientific, medical, and practitioner communities. These community partners have central positions and extensive activities, which contribute to the widespread dissemination. Dissemination allows the focal firm to gain feedback and knowledge spillovers about its products and competitors, which can help adjust the firm’s innovation plan and position (Figure 1(d)). Focal firms often chose to disseminate the information about their projects to three types of partners: (1) Key Opinion Leaders (KOL), such as physicians who influence their peers; (2) academic researchers, who use the focal firm’s compounds and related technical support through the Material Transfer Agreement (MTA); and (3) venture capital investors, who communicate the focal firm’s projects to a broader network of investors and potential partners for later stages.

4.2. Multiplicity of learning: leveraging partner variety

The findings show that early-stage entrepreneurial firms, despite constrained resources and incipient capabilities in alliance management, engage with various partners using multiple learning mechanisms, progressively. Over time, the focal firm expands its alliance network from one to multiple partners, and its learning from one to several types of knowledge, exploring new knowledge and new collaborators for innovation and growth (Figure 2).

Early-stage firms may take different routes to gain diverse types of knowledge from multiple partners, and the multiplex learning from one partner is only an intermediate step in one route. With evolving knowledge needs and learning opportunities discovered, focal firms leverage new interactions or new partnerships to apply the accessing, transferring, catalyzing, or disseminating mechanisms of knowledge building. Careful early steps in the route making up the weak network position, focal firms achieve rich alliance learning extended from existing ties and the use of multiple mechanisms over time (Figure 2, Route 1 to 5, Route 2, and Route 3 to 4).

The case of Firm Alpha exemplifies the development of its preclinical skill starting with a technology licensing collaboration with large pharma HD for the lead optimization stage (Table 3, CoL3). During the contracted lead compound knowledge transfer, Firm Alpha learned about HD’s preclinical supplier network and managerial approach (Figure 2, Route 1). Then, this knowledge directly fostered the creation of new collaborations with other partners, for instance, the preclinical CROs, for the preclinical stage learning of multiple knowledge types (Figure 2, Middle route 5, toward upper right). Firm Alpha then consulted with multiple preclinical CROs (Table 3, CoL4) for a specific preclinical design (Figure 2, Left route, step 3). Subsequently, Firm Alpha also gained the CRO managerial knowhow from its experience of working with multiple CROs (Left route, step 4). Ultimately, using multiple mechanisms, Firm Alpha gained and leveraged knowledge from several collaborations to match its internal R&D activities and, in turn, developed a complete preclinical capability, ranging from technological knowhow, network resources, to preclinical study design and supplier management. This diverse knowledge was then redeployed to the preclinical studies of four other drug candidates of the company. Similarly, Firm Alpha adopted this strategy for its clinical skill building (Table 3, CoL5 and CoL6). These examples indicate how early-stage firms combine several partnerships to provide diverse knowledge outcomes through multiple learning mechanisms.

Table 3. Multiplicity of learning in early-stage entrepreneurial focal firms: examples.

Routes to multiplicity	Multiple partners	Multiple learning mechanisms		Total knowledge accumulated
		Initial	Subsequent
From one partner, one type of knowledge to multiple partners, multiple types of knowledge	CoL3 (in-licensing partner, a large pharma) followed by CoL4 (multiple CROs)	Lead stage: Contracted knowledge transfer	Preclinical Stage: new knowledge accessing, transferring, and catalyzing	Lead technology Supplier network information Supplier management knowhow Preclinical study knowhow CRO management
	CoL5 (CRUK), CROs and teaching hospitals followed by CoL6 (HGS) and other clinical CROs, teaching hospitals, etc.	Clinical stage for 1st product: Contracted knowledge accessing	Clinical stage for 2nd product: new knowledge accessing, transferring, and catalyzing	New clinical data Clinical study new insights Product feedback & market info
	CoL25 (BioS) followed by CoL26 (BRI) and CROs	Target/lead stages: Contracted knowledge accessing	Lead/preclinical stages: catalyzing, new transferring, and more catalyzing	Lead technology Preclinical study knowhow New partner info
One type of knowledge learned and enriched through multiple partners and learning mechanisms	CoL11 (a group of universities)	Contracted (technology) transfer	More transferring, and catalyzing (normally at the same stage)	New insights, new directions, or new ways of filling its knowledge gaps: better protocols, methods, and compound alternations
	CoL13 (some 80 scientists and researchers through MTA)	Contracted knowledge accessing	Disseminating and catalyzing	Discovery of critical errors and opportunities for exploration of new animal models, indications, and chemical combinations for the drug candidates

Note. CoL: a collaboration among the sampled 26 collaborations; CRO: Contractual Research Organization (preclinical and clinical); MTA: Material Transfer Agreement. Source: Authors’ elaboration.

In addition, an important form of learning multiplicity is identified when the early-stage firms enrich one type of knowledge through multiple partnerships, which then push the learning toward the knowledge variety (Figure 2, Route 3 to 4). This route allows focal firms to use multiple collaborations for the exploration of new insights, directions, or ways of filling knowledge gaps at a specific stage of product development. Step 3 often involved various research subcontractors such as universities and research centers. Taking advantage of the low cost (compared with in-house or high-stake options) and relatively competitive market, subcontracting specific research tasks in early stages economized the deployment of limited resources in focal firms, while it frequently generated novel technological knowledge about the project, such as new candidate compound applications, or adjustments of the product development trajectory to address unforeseen errors. Intensive experiments, expensive equipment, and a broad spectrum of specialties in universities and research centers helped facilitate this crucial knowledge benefits beyond the initial contractual purpose, meeting the exploratory needs in the focal firm’s knowledge building.

Firm Charlie, for example, collaborated with multiple universities for testing an HIV drug compound in development. It engaged in regular and novel trial designs for better protocols, methods, and compound alternations from a number of collaborating universities (Table 3, CoL11). Working with frontier university researchers and labs, the early-stage focal firm leveraged the relationship to catalyze its R&D knowledge building for lead, preclinical, or clinical stages beyond the initial tasks and knowledge transferred (Figure 2, Route 3). During the collaboration, Firm Charlie gained more than the testing results from the university partners. As a result, it uncovered new directions in drug development. In another example, Firm Delta disseminated its compound to a global network of some 80 scientists and researchers for their research in order to gain and catalyze new knowledge and insights about its product and the testing process (Table 3 CoL13, Figure 2 Routes 3 or 2).

4.3. Organizational goals, entrepreneurial actions, and emergent learning

When examining the logic underlying the multiplicity of learning strategies, the data indicates a comprehensive view of the focal firms’ partnership process. This process starts with the organizational goal of an early-stage firm (Figure 4). The goal guides key decisions and essential entrepreneurial actions for learning, such as identifying learning opportunities, and selecting and engaging partners to develop new knowledge (Figure 4). Multiplicity of learning is therefore an emergent process situated in the firm’s specific strategies, developmental stages, and evolving resources and interactions with new partners. To illustrate this process, two representative early-stage firm cases (Firm Alpha and Firm Golf) are presented below. The two cases differ in characteristics but converge in manifesting the concept of multiplicity. Firms Alpha and Golf were in different therapeutic drug development stages during this study period, leading to their distinct partners, resources and capabilities (Figure 3).

Figure 4. Multiplicity of learning process in alliance-networks: an integrated model.

Source: Authors’ elaboration.

4.3.1. Human therapy biotech firm Alpha

Created from the merger of two university spinoffs, Firm Alpha inherited the discovery stage expertise and started several lead optimization programs (to discover drug candidates) at the founding with about 32 chemists and biologists. Ten years after the merger, it had developed three proprietary new drug programs at the preclinical and various clinical stages respectively. Firm Alpha’s partners included investors, scientific and clinical leaders, universities for research and discovery, as well as development collaborators (hospitals, large pharmas, and other potential licensees for the clinical stage and beyond).

Organizational goals, knowledge gaps, and partner identification. Firm Alpha had a strategy of becoming a dedicated full-line biotech company in the new therapeutic market. With over 100 patents, Firm Alpha focused on neurodegeneration disease therapies, knowing that it had to pursue the developmental milestones and build the required capabilities for the goal (Figure 3). It achieved fast growth in its early years, fueled by multiple venture capital funding and supported by its various partnerships. Firm Alpha moved from the preclinical stage into the clinical stage facing a dramatic knowledge gap. It needed to seek external help in clinical design, testing knowhow, and marketing expertise.

Initially, Firm Alpha used smaller partnerships to initiate the clinical study for a less-important drug candidate in its pipeline, accumulating knowledge for the developmental stage. The initial partnerships included eight clinical CROs, two teaching hospitals, and a non-profit research institute in the U.K. (CRUK). Subsequently, Firm Alpha sought an equity-based alliance with a large biotech pharma, HGS, for its major drug candidate’s clinical stage development. This major partnership formed after a four-month search that identified a match with HGS given the latter’s multiple clinical trial experience and programs. The non-equity based early partners contributed to Firm Alpha’s articulating the direction and needs for the equity-based alliance. Additionally, the CRUK experience allowed Firm Alpha’s scientists to move their learning curve quickly toward the clinical CRO, while the teaching hospitals further fostered the learning as the ‘test drive’ in some self-monitored clinical trials. According to the COO of the firm:

…they have been able to help us in saying which tests are relevant and useful and which tests are not relevant and useful. And so we went in there with some preconceived ideas about what is important to do and we were completely wrong…

Therefore, Firm Alpha’s clinical stage knowledge accumulation allowed it to ‘make the most correct decisions to the project’. The important HGS alliance built on Firm Alpha’s earlier multiplicity of learning, which then fostered increased learning from HGS and other partners (Table 3, CoL5 and CoL6).

Partner engaging. According to the managers in Firm Alpha, ‘collaboration with large pharmas often does not automatically lead to knowledge learning for small biotechs’. Firm Alpha promoted multi-level interfirm interactions in the contracted drug development project to engage the partner and achieve the desired learning. To allow for the timely data and progress sharing, Firm Alpha added an intermediate level of meetings between the initially set two levels of interaction, namely, the Joint Research Committee (JRC) that oversaw the overall collaboration and met only every three months, and the operational team inside Firm Alpha that produced data on the project stored for the JRC meetings. The intermediate-level interaction allowed managers and scientists to call meetings when needed, fostering Firm Alpha’s learning.

The intermediate level meetings really engaged HGS scientists and managers in teaching and answering questions so that Firm Alpha learned to ‘know what’s important and what’s not important, what pushes the right patterns and what doesn’t’. This initial learning in technology quickly expanded to multiple types of knowledge, such as IP strategies, risk management, and quality insurance. After HGS, Firm Alpha largely completed its clinical knowledge accumulation except for small gaps. These knowledge needs were then addressed via CROs, teaching hospitals, or specific licensing in successive partnerships. Firm Alpha continuously used the model of engaging small initial partnerships followed by larger strategic partners for multiplicity of learning. The COO stated:

…you learned, you evolved and so when we went to [the next partner as] we were, we know exactly what they want to hear, and we told them it was true.

4.3.2. Human therapy biotech firm Golf

The case of Firm Golf is a reinvented entity with key genomics assets including technology and personnel from a failed bio-pharma company. With government funding, Firm Golf grew to 30 employees and developed its key technological asset, an innovative platform for therapeutic target discovering, the ‘SR’ platform. Three years after its founding, the firm downsized to 10 employees given financial difficulties, setting a new strategy of ‘establishing into a clinical-stage biotherapeutic company’, expecting a high-value acquisition in the future.

Organizational goals, knowledge gap, and partner identification. Management decided to focus on developing therapeutic monoclonal antibodies (mAb), a promising technological novelty in new diagnostic tests and therapeutics design (Robbins-Roth, 2000). The firm’s assets included the leading ‘SR’ platform technology in target discovery and a portfolio of proprietary therapeutic targets identified by the platform. Along the revenues generated from its SR platform and targets, Firm Golf aimed to build expertise in mAb technology to implement its therapeutic mAb strategy departing from this target stage (Figure 3).

To achieve this goal, Firm Golf kept strengthening its ‘SR’ platform technology through existing partnerships with universities, hospitals, and frontier technology licensing from other biotech companies. In addition, it developed three proprietary focused target programs, such as oncology therapy, and expected to accelerate its preclinical studies by leveraging other partners’ mAb expertise, including bringing lead compounds and conducting testing for those compounds.

Given its knowledge gap in mAb technologies, Firm Golf sought new partnerships to develop novel mAb drugs against its proprietary targets for disease treatment. To search for partners, the firm focused on:

…attending scientific meetings and, therefore, establishing a very strong collaborative oncology network with university hospital centers and scientists providing access to clinical material and insight…

It selected various partnerships to gain knowledge related to the mAb drug development. Firm Golf licensed-in a small molecule (not mAb) drug compound, for example, to gain experience of preclinical studies and the IND regulatory process. Small molecule drugs are relatively simple chemical compounds, ideal to help Firm Golf learn about drug development beyond the first target discovery stage. This knowledge turned out important preparation for searching the future mAb partnerships.

Partner engaging. Firm Golf managers stated that its earlier partnership experience helped find an ‘ideal’ alliance partner for its strategic learning purpose (CoL25). Consequently, it partnered with BioS, a large diagnostics company with strong mAb expertise in the ovarian cancer area. BioS provided mAbs (compounds and knowhow) for the therapeutic R&D of Firm Golf. In exchange, BioS would use Firm Golf’s proprietary targets and the drug candidates to be developed during the partnership to then create diagnostic kits. The partnership allowed Firm Golf to access, and transfer some of, BioS’s mAb technologies and knowhow. BioS also manufactured all the mAbs Firm Golf needs during the latter’s R&D.

Firm Golf’s learning multiplicity went beyond the accessing and transfer, involving BioS and its network (Figure 2, routes 1 and 2), as well as subsequent partnerships (Figure 2, route 3). Firm Golf developed an ‘excellent relationship’ with BioS, and then expanded this collaboration into more targets and mAbs, new therapeutic areas, and wider access to large-scale R&D facilities and technologies:

…before we started this partnership, we hadn’t really worked on antibodies and so this partnership allowed us now to … develop new expertise and increased the level of knowledge around working with these antibodies …to go into animal models and to generate other partnerships … they teach us a lot …

Firm Golf actively promoted frequent information exchange through mutual visits, multi-department in-person and teleconference meetings, numerous emails and phone calls, and informal meetings in scientific conferences. In the words of Firm Golf CEO:

…you don't hesitate to go as far as possible in [devoting] important information to help your partner make its informed decision, because the more you invest there, the more lucrative will be your deal. Essentially, that’s very key learning. So … we were told by them that we are for the first time they had access to very, very valuable scientific information in making their decision and their selection of targets…

The evolving relationship nurtured a perception of a common objective in cancer diagnosis and cancer treatment that facilitated multiplicity of learning.

Moreover, the BioS partnership allowed Firm Golf to become part of a bigger network including large pharma companies and the mAb community. It raised Firm Golf’s reputation for future partnerships with other organizations, expanding its innovative programs. For example, the collaboration with a national research institute (Table 3, CoL26) occurred as a ‘hand-in-hand deal’ with the BioS partnership (Table 3, CoL25). The institute used its specialized humanization technique to advance the ovarian cancer mAb compounds from the BioS partnership’s preclinical studies to the clinical studies stage. Important exchange of science and knowledge between the partners, as well as more collaboration in new programs and activities such as personnel training contributed to the expansion of the BioS alliance into more areas.

Deliberate partner engaging actions also helped Firm Golf enhance its long-term relationship with universities and the research community, for advancing the discovery platform and subsequent developmental stages (Figure 3). A Firm Golf manager stated:

Well, it’s the technology platform; it’s also the network of collaborators…Some of these people are aware of the progress that we are making and will in turn support us for the design of the clinic trials and eventually have access to patients.

The multiplicity of learning with deliberate use of partnerships transformed Firm Golf from a target discovery platform company to a new drug development company in niche therapeutic areas. Actively identifying partnerships and the emergent leveraging actions facilitated its entrepreneurial value creation.

5. Discussion

5.1. An integrated model of multiplicity of learning in alliance-networks

Based on the findings, this study advances a model of startup firms pursuing multiplicity of learning in alliance-networks to support their entrepreneurial development. Through the in-depth case study method, the current study establishes this new concept that captures an integrated process model of how learning in alliance happens in a resource-constrained context for innovation and growth (Figure 4). In this integrated model, the young firm achieves the ultimate knowledge diversity through various routes, using multiple partnerships progressively, each with varied knowledge. This model goes beyond the ‘multiplexity’ strategy that focuses on varied knowledge from a single partner (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018; Lowik et al. 2012; Moghaddam, Bosse, and Provance 2016). ‘Multiplexity’ is recognized here as an intermediate step in one route of the overall alliance learning strategy in entrepreneurial firms aiming for knowledge variety through partner variety (Figure 2).

The integrated model reconciles the argument of multiplexity (one partner) and the prevailing partner variety in the entrepreneurial alliance learning literature by incorporating the role of organizational goals (Macpherson and Holt 2007; Martínez-Noya and Narula 2018), including both the overall business goal and the staged project goals. The early-stage biotech firms in this study reported at least four partners on average, and learning from various partners played a crucial role in their growth. Partnerships both are influenced and influencing the strategic direction in early-stage firms, filling the evolving learning gaps, and meeting the diverse needs in firm innovation and growth. The sole emphasis on learning efficiency may validate the multiplexity argument; however, taking account of various organizational goals indicates a more comprehensive approach necessary to understand the actual learning strategies adopted in firms with resource constraints. Consistent with the startup innovation literature, exploratory alliances formed with various partners are an important part of the learning strategy in startups for diverse knowledge needed for innovation and growth (Gronum, Verreynne, and Kastelle 2012; Lavie, Stettner, and Tushman 2010; Leiponen and Helfat 2010; Simba 2015).

Furthermore, by taking into account the staged project goals for product development, this model provides a dynamic perspective of alliance learning in the startup’s entrepreneurial process (Macpherson and Holt 2007; Muñiz Ávila, Silveyra León, and Segarra Pérez 2019; Pisano 2006). The ongoing entrepreneurial development constantly reveals new knowledge gaps and learning opportunities, and therefore informs the actions regarding who to collaborate with and how to engage partners. After a round of knowledge gaps filled and specific goals achieved, the new goal guides the next round of learning and development. The progressive learning process leads to adopting a multiplicity of learning strategies to gain various types of knowledge, which appears both deliberate and emergent: it is goal oriented along the drug NPD process on one hand; and emergent on how partners and the four mechanisms are engaged with growing resources and capabilities on the other hand.

Choosing and combining partnerships and learning mechanisms represents an entrepreneurial action that effectuates the alliance-network learning strategy under conditions of limited resources (Sarasvathy 2001). This increases understanding of the interrelations between learning strategies and partner selection as well as the engagement actions, both aspects identified as needing more research in the R&D alliance literature (Martínez-Noya and Narula 2018). These early-stage firms purposefully connect their partnership strategy with the immediate and future knowledge needs, despite having different starting points and strategic trajectories.

Specifically, this integrated model focuses on the unique entrepreneurial context and highlights the young entrepreneurial firm’s priority on integrating learning mechanisms for innovation instead of efficiency (Figueiredo, Larsen, and Hansen 2020; Kamuriwo and Baden-Fuller 2016; Powell, Koput, and Smith-Doerr 1996; Simba 2015; Simba and Ndlovu 2014). Advancing existing understanding about collaborations’ effects on learning in small and innovation-oriented firms, this study extends recent work on the integrative learning strategies of biotech multinationals for innovation capability building (Figueiredo, Larsen, and Hansen 2020). Early-stage entrepreneurial firm contexts can differ from other contexts such as small businesses not focused on creating new products, or multinationals examined in Figuereido’s study that uncovers the deliberate interactive learning with local subsidiaries and other organizations (Figueiredo, Larsen, and Hansen 2020). Our study adds to these research streams an integration of alliance learning strategies adopted by the entrepreneurial start-ups to build a wide range of new knowledge through various partnerships. For example, the firms actively seek a large number exploratory partnerships with long-term relationship for innovation. Using firm-level and collaboration-level data, as well as longitudinal data along the NPD stages and the early stages of an entrepreneurial firm’s development, this study fills the identified dearth of more detailed studies on learning strategies from an integrative and firm level perspective (Figueiredo, Larsen, and Hansen 2020). This study contributes to understanding how the young firm and its alliance evolve over time, expanding the organization-specific process approach to knowledge acquisition and learning (Cope 2005; Figueiredo, Larsen, and Hansen 2020; Macpherson and Holt 2007; Martínez-Noya and Narula 2018). The articulated organizational process involving deliberate partner identifying and engaging practices in alliance-network learning (Figure 4) adds an important analysis that focuses on the entrepreneurial firm’s strategic purpose, rather than the collaboration goal.

5.2. Multiplicity of learning: network resourcing in entrepreneurial firms

The multiplicity of learning mechanisms with various partners demonstrated in this study supports and extends existing literature on the learning complexity for entrepreneurial firms and uncovers how the learning happens (Kamuriwo and Baden-Fuller 2016; Macpherson and Holt 2007; Martínez-Noya and Narula 2018; Muñiz Ávila, Silveyra León, and Segarra Pérez 2019). From a perspective of dynamic knowledge sourcing in networks, or ‘network resourcing’, this study highlights the emergent learning process after knowledge transfer and accessing from partners, and offers a new lens in understanding this alliance learning complexity (Feldman and Worline 2011). The literature on alliance learning highlights knowledge transfer and access given existing knowledge in a tie (Grant and Baden-Fuller 2004; Inkpen and Tsang 2007; Rothaermel 2001). Existing research on network resources emphasizes the information about choice of partners, alliances, opportunities for new ties and partnerships, and gains in reputation and funding (Gulati 2007). This study goes beyond existing work by showing that entrepreneurial firms influence and shape the information and knowledge resources embedded in a tie or network and, in turn, shape the learning outcomes they achieve. Catalyzing and disseminating are more active mechanisms than transfer and access in its interaction with partners. Actively and selectively adopting and transforming the external knowledge available through its network of collaborators, an entrepreneurial firm can achieve better learning than those who don’t.

Firms take purposeful and structured actions in preparing learning events, and use entrepreneurial approaches in resource acquisition and deployment. They may take risks, for example, giving partners opportunity and time to bring exploratory value while tolerating short-term failure. In addition, they are proactive, for example, planning for multiplexity and multiplicity as early as in partner searching process while attempting to use the emergent spillovers. Moreover, they are innovative in discovering new partners or spillovers from collaborators, and in deploying creative ways to enhance relationship and generate synergies. Network information and knowledge benefits can be gained without necessarily repeating ties with same partners as existing literature argues. Instead, crucial knowledge emerges through purposeful and multiple actions in different ties with various partners.

Therefore, this study captures a new dimension in entrepreneurial firm developmental heterogeneity due to the different processes in the firm’s knowledge sourcing. The existing literature typically assumes varied knowledge sourced from a variety of types of organizations or ties (that is, supplier, government, customer) for innovation (Baum, Calabrese, and Silverman 2000; Gulati 2007; Leiponen and Helfat 2010). Nonetheless, entrepreneurial firms can set different organizational goals, recognize different knowledge gaps, and make different decisions regarding partner selection and engaging (Figure 4). Heterogenous actions in using the multiple learning mechanisms and their combinations from various partners can explain important variance in organizational consequences such as capability development and innovation output.

As a specific aspect of network resourcing, this study also offers new insights on young entrepreneurial firms overcoming two main learning challenges. One challenge the literature emphasizes is lacking prior alliance knowledge for learning (Bojica, Estrada, and del Mar Fuentes-Fuentes 2018; Yoo, Sawyerr, and Tan 2016). This study demonstrates that entrepreneurial firms deliberately prepare for major alliance learning through smaller and non-equity partnerships. Through this purposeful but exploratory network, they build ‘prior’ knowledge in both technology and relationship for further significant learning. This differs from what is typically argued for prior knowledge acquisition through internal investment. According to the literature, a second challenge is the young firm’s incipient network position and low relational capability. This study demonstrates that entrepreneurial firms substitute the established position or capability with deliberate and substantive efforts in partner engaging practices. Extensive engaging allows for uncovering mutual interests and additional learning opportunities from the same collaborator. They also foster more collaborations within and beyond the current partner’s network guided by the organizational goals. They increase knowledge about the collaborator, partnership management, and the overall R&D process, an exploration that justifies a long-term relationship for innovation. Through multiplicity, the early-stage firm actively enlarges its network by searching for new partners. This study also supports and extends work on network resources showing these young small firms leverage their alliance network to gain ‘network resources’, that is, information regarding further partnership and learning (Gulati 2007).

The network resourcing perspective adds a firm-level understanding on knowledge circulation and the dynamism of city regions (Gertler and Vinodrai 2009; Niosi and Bas 2003; Schiffauerova and Beaudry 2012; Tremblay and Klein 2014). Existing studies in Canada document the growth of knowledge-based clusters such as the Montreal biotech one. Some emphasize the role of policies toward high-tech and innovation. Others that analyze geographic proximity explain that just sharing the same knowledge base is insufficient. The new findings presented highlight the firm strategies to leverage knowledge variety within and beyond the cluster providing a different dimension than physical proximity. In Canadian biotech clusters with a majority of small start-ups, the model of multiplicity of learning provides an explanation of how these firms contribute to new knowledge generation and, in turn, cluster growth. Groups of such entrepreneurial firms achieve this by dynamically developing their own capabilities in navigating the evolving alliance-networks and learning.

6. Conclusion

This study contributes to entrepreneurial development, small firms, and alliance learning research by proposing the new concept multiplicity of learning, based on a model that emphasizes an integrative organizational approach. Examining the question of how early-stage entrepreneurial firms strategize to manage learning with multiple partners, this perspective extends research on the effects and interaction between constrained resources and the entrepreneurial firm learning in alliance. It shifts the current theoretical framework focused on economic use of existing resources to one that highlights the process of ongoing network creation and knowledge resourcing. It also advances a view of more deliberate deployment of entrepreneurial learning strategies and capability building mechanisms connected to the firms’ organizational goals. Considering both the entrepreneurial development process and the innovation process, this study advances the literature of alliance learning by illustrating the strategies of early-stage entrepreneurial firms. Selecting their strategies in acquiring and gaining new knowledge with alliance partners, these firms build their networks to meet the development goals. Understanding the multiplicity of learning strategy recognizes the possibility of effective multi-partner learning in early-stage entrepreneurial firms by continuously and deliberately discovering learning opportunities and using suitable mechanisms to learn from various partners. It extends the current literature scope to a more comprehensive strategic approach including various entrepreneurial firms with different goals. This multiplicity strategy approach, therefore, facilitates both the broader access to and gains from varied knowledge sources associated with varied learning mechanisms. These firm-specific learning activities, guided by the organizational goals and emerging around the evolving inter-organizational collaborations, drive the significant focal firm knowledge benefits in alliance networks. This also deepens understanding of the small-firm dominated Canadian context where the multiplicity of learning in alliances may be contributing to its dynamism.

To business practitioners, this study offers important insights regarding the strategic choices of knowledge acquisition facing small early-stage firms under constrained resources. Taking into consideration the organizational goal, a young firm may still adopt an entrepreneurial strategy exploring several alliances for learning multiplicity, rather than being limited to one alliance. Resources and managerial challenges in multiplicity strategy can be addressed by progressive alliance-network building. As well, substantive and evolving firm and inter-organizational efforts unfold in the process of diverse, emergent learning. This process is entrepreneurial in nature, indicating an upgraded skill set required for the managers. To gain the knowledge variety needed to innovate, effective management involves proactive efforts to connect goals, recognize gaps, identify and engage partners, and use a variety of learning mechanisms to build innovation capabilities.

To industry policy makers, this study has practical relevance today for supporting entrepreneurial firms, in general, to pursue effective alliance learning with a variety of partners. Resource constraint urgency, such as that brought by the COVID-19 pandemic, significantly impacts the biotech industry, both negatively and positively. On the negative side, as in many industries, biotech companies have seen their R&D activities adversely affected (BIOTECanada 2020). At the same time, the current pandemic creates demands and opportunities for early-stage companies across multiple industries to engage in more collaborations in developing diagnostic and therapeutic solutions (BIOTECanada 2020). The issues this study addresses relate to the competitive and innovative success needed to respond to the current pandemic, or similar challenges in the future in health, social, and environmental domains, where biotech firms pursue new solutions. As governments, firms, universities, and others engage in health innovation and economic recovery, this study offers insights on the role of effective learning and diverse partnerships to advance innovation and growth at the firm level, and the importance of supporting the development of early-stage small companies.

Several limitations of the study open opportunities for future research. Given the focus on one country, future studies could explore cases in different locations for comparison. In addition, this study does not examine the complexities of young firms operating in global markets, which is a trend in the innovation field. Future studies could add the dimension of complexity in managing international collaborators to validate the proposed model and further enrich understanding on learning multiplicity. Finally, a way to measure resource constraints may be examined in theory testing studies in the future, as some startups may have more resources and stronger network position than others. This could expand a nuanced understanding of the relationship between resource constraints and alliance learning strategies.

Disclosure statement

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

Additional information

Notes on contributors

Yuanyuan Wu

Dr. Yuayuan Wu, is Associate Professor of Entrepreneurship and Strategy in the Faculty of Business Administration at Lakehead University (Orillia Campus), Ontario, Canada. Her research focuses on the strategies and processes of capability and enterprise development. She studies these topics in the contexts of the Canadian innovation-based industries and the emerging high-tech sectors in China, including a Social Sciences and Humanities Research Council of Canada (SSHRC) funded project on the aerospace industry collaboration between Canada and China. Her other projects involve the areas of corporate entrepreneurship and platform ecosystem dynamics; she recently co-authors the book “Disruptive Innovation through Digital Transformation: Multi-Sided Platforms of E-Health in China.” Professor Wu holds a Ph.D. in Business Administration from McGill University.

Paola Perez-Aleman

Dr. Paola Perez-Aleman, is Associate Professor of Strategy and Organization in the Desautels Faculty of Management at McGill University. The focus of her research is on the processes that build capabilities of enterprises, foster innovation, and advance sustainable development. Specifically, Dr. Perez-Aleman examines how clusters and inter-firm networks emerge and grow; how enterprises in developing and emerging economies build innovation capabilities; and how small and medium enterprises in global value chains pivot to sustainable production. Her research explores organizational and institutional processes that foster equity, innovation, and sustainability. Professor Perez-Aleman holds a Ph.D. from the Massachusetts Institute of Technology and a B.Sc. from the University of California at Berkeley.