Abstract

Affordances and generativity of digital technologies have led to a proliferation of digital entrepreneurship, when new ventures use digital artifacts as the core component of their value creation and appropriation strategy. Digital technologies that are open-sourced are the foundations of the modern digital economy. Open-source collaboration (OSC) platforms, e.g., GitHub, have transformed the digital entrepreneurship ecosystem, with 15 of the top 20 unicorns having GitHub repositories. Lin & Maruping (2022) study the causal impact of the types of open-source collaboration on venture outcomes in a quasi-experimental setting, where digital new ventures participating in open-source collaboration through GitHub public repositories are considered as a treated group, with new ventures without public repositories as a control group. One of the key omitted variables in the study was the new venture business model and its novelty. Guzman & Li (2023) have proposed a measure of new venture differentiation using natural language processing of new venture business models and examined its relationship to venture outcomes. We build on these two studies to examine the nuanced dynamics of open-source collaboration, new venture novelty, and venture outcomes. Our key research question is how open-source collaboration affects the relationship between new venture novelty and venture outcome. Additionally, we examine how this effect varies across venture stages. We address this research question by introducing a new measure of venture novelty, leveraging developments in large language modelling to create a semantic competition network of new ventures (Aceves & Evans, 2024) and adopting information information-theoretic approach to operationalize novelty (Aral & Dhillon, 2023). Our key hypothesis is that inbound OSC has a more positive moderating effect than the outbound OSC in the venture conception stage, while the outbound OSC has a more positive moderating effect than the inbound OSC in the venture growth stage. The estimation strategy uses a weighted least squares (WLS) approach with firm and time fixed effects with inverse probability of treatment weighting (IPTW) for mitigating the selection bias.

Comments

tpp1434

Download

Share

COinS