Answers are grounded in this paper and cite their sources.
Reward-based crowdfunding has become an increasingly relevant mechanism for financing and testing innovative projects, yet success rates vary considerably across categories. This study introduces a structured multi-method approach that combines short-text clustering, regularized inference, and interpretable prediction to examine the drivers of crowdfunding performance on Kickstarter, while ensuring transparency in how results are generated and understood. Applying the framework to over 300,000 Kickstarter projects across the top ten categories, we integrate Group Lasso regression, Gibbs Sampling Dirichlet Mixture Model, decision trees, and Generalized Additive Models to systematically examine how project characteristics, thematic content, and temporal factors influence funding outcomes. The study reveals category-specific success factors and non-linear relationships between various predictors and crowdfunding success. Key findings highlight the importance of dynamically addressing project themes, launch timing, and funding goals, with effects varying across categories. Additionally, the study introduces a research method that enables entrepreneurs to optimize their market position. Overall, our research advances theoretical understanding and provides readily applicable insights into the interplay of crowdfunding temporal dynamics, thematic content and project characteristics.
No takes yet. Share an insight, caveat, or question.
Songhori et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: