Digital identities of female founders and crowdfunding performance: an exploration based on the LDA topic model

Purpose This paper aims to investigate female founders' digital identities (i.e. the founder's online self-representation) and explore how female founders' digital identities affect crowdfunding performance. Design/methodology/approach Leveraging a data set of 3,125 Kickstarter crowdfunding campaigns launched by women between 2014 and 2019, this study uses the latent Dirichlet allocation (LDA) topic model to investigate female founders' digital identities and explore their impacts on crowdfunding performance via logistic and Tobit regression analyses. Findings This study finds that on crowdfunding platforms, female founders display four main types of digital identities: educational practitioner, experienced entrepreneur, creative innovator and life dreamer. Moreover, our results reveal the differential influences of these identities on crowdfunding performance. Specifically, the experienced entrepreneur or educational practitioner portrayal may benefit fundraising; the life dreamer portrayal may hinder it; and the creative innovator portrayal may have no significant effect. Practical implications This study suggests how women can use their digital identities to improve their crowdfunding performance, indicating a fruitful way to overcome female entrepreneurial financing hurdles in the digital age. Originality/value This study highlights the digital identities of female founders engaged in crowdfunding, enriching the emerging research on digital identity in entrepreneurship and crowdfunding. Furthermore, the application of the LDA topic model in the study enlightens future research on leveraging big data analytic methods to examine female entrepreneurship in the digital economy context.