A Tale of Two: How Network Agency Influences Network Creation for Start-ups Chiamaka Kwazu Nottingham Business School, Nottingham Trent University Chiamaka.kwazu@ntu.ac.uk Abstract: There is an existing consensus that networks positively impact start-ups. New ventures gain access to market information, funding, emotional support and improved legitimacy from networking. However, existing efforts at articulating network creation and impact on new ventures are skewed due to a strong focus on structural properties. Although these properties contribute to understanding network creation and impact, they often ignore the individual who is either a network beneficiary or broker. This study examines the role of the network actor’s agency in entrepreneurial network creation. Understanding agency is critical for explaining how individuals make sense of network opportunities available within their context. Network agency reveals how entrepreneurial networks are created, what influences actor motives and behaviour, and the brokerage methods adopted to meet these motives. This study explores network agency in two contexts: the interventionist incubator context and the location-induced cluster context. The Lagos Technology Ecosystem comprised of different incubator models and cluster is examined. 31 semi- structured interviews with start-ups from both contexts were conducted to get a nuanced view of how the context influences entrepreneurs' network agency. This paper reveals that network motives like personality, reciprocity, and legitimacy as the main motives for tenant firms to engage in network activities. Conversely, cluster firms listed trust, and the need to access knowledge, information and opportunity as key reasons for engaging in network action. In both contexts, direct and indirect brokerage is utilised to meet the motives triggered. However, the dynamics and nature of these brokerage methods are different across the two contexts. Tenant firms demonstrate a tertius Iugnes behaviour, while cluster firms demonstrate separation behaviour. Keywords: Social Network, Agency, Entrepreneurial Networking, Networks, Lagos 1. Entrepreneurial Network and Networking Impact Networks act as stimulants to the entrepreneurial process, as it allows entrepreneurs to obtain information and resources needed for battling business uncertainties (Chell and Baines, 2000). Additionally, Witt (2007) views that existing network studies hypothesize a positive relationship between networking and entrepreneurs. Therefore, there is a need to account for why individuals and groups enact certain networks. This will account for how and why the structure and processes of embeddedness affect entrepreneurs, and how this contributes to variations in the form of entrepreneurship generated (Slotte-Kock and Coviello, 2010; Mckveer, Anderson and Jack, 2014). Recognising the importance of these relationships in shaping organizational behaviour and firm performance, studies seem conflicted on how individuals make decisions; whether it is done independent of network structures or embedded within the dynamics of the network structure (Tasselli and Kilduff, 2019). This paper joins the emerging network agency discussion, exploring how individuals leverage network agency in creating and making decisions independent of their network structures. Agentic tools like network motivation and brokerage are used to understand how entrepreneurial networking is created within the entrepreneurial context of business incubators and enterprise clusters. Studies like Kidluff and Brass (2010) and Tasselli et al. (2015) argue that the exclusion of individuals from social network research has never made sense, as researchers tend to treat network structures as given, and therefore pay less attention to how actors create, perpetuate and modify network structures through their actions (Gulati and Srivastava, 2012). According to Tasselli and Kidluff (2021), relying on the structural approach to networking and network creation ignores the importance of individual volition and action in network processes. Therefore, Emirbayer and Mische (1998) argue that empirical network studies continue to privilege network structure over network action initiators. They add that even when the individual's role is noted, it is often under-theorized. Equally, structuralist perspectives on network creation and impact have yielded conflicting findings, as attempts have been made to validate the impact of structural holes and closure; see studies like Kidluff and Oh (2006); Semrau and Werner (2013); Redlich et al. (2013) and Kreiser (2011). A plausible explanation for these varied views is provided by Jack et al. (2010) study, which opines that networking and networks exist as dynamic endeavours and relationships that could induce various changes within the network creation process. Hence, compositional ties that might be useful at one point might not be later. This focus on network structural attributes relegates the individuals at the centre of the network creation process to passive agents, even though the networks created, and the resulting network structures are functions of individual choices and needs (Willer 509 Proceedings of the 18th European Conference on Innovation and Entrepreneurship, ECIE 2023