Marketing for Early-Stage Funding

This dissertation examines how entrepreneurial communications shape funding outcomes across the two stages of reward-based crowdfunding: the pre-goal stage, in which entrepreneurs seek to reach the all-or-nothing goal, and the post-goal stage, in which they continue raising funds after crossing that. Because Kickstarter follows an all-or-nothing structure, the meaning and effectiveness of entrepreneurial communications vary across stages and depend on whether a project is still pursuing its goal or has already achieved it and is seeking additional funds. Across both essays, I show that understanding crowdfunding requires explicit attention to these pre- and post-goal stages. Essay 1 asks when updates are most effective during a crowdfunding campaign. Answering this question requires a framework that accounts for the dynamic effects of updates, as well as the endogenous nature of the provision point. Entrepreneurs post updates strategically in response to evolving campaign conditions, goals are themselves chosen endogenously, and the effects of updates evolve nonlinearly over time. To address these challenges, I develop a Bayesian provision point model that builds on the insight that entrepreneurs benchmark their decisions against similar past campaigns. This benchmarking provides exogenous variation that helps instrument for the non-random timing of updates and correct for endogenous goal-setting. I also incorporate Gaussian Process priors to capture smooth, stage-specific dynamics in update effects. I find that updates have a modest positive effect before the goal is reached, but their effectiveness becomes strongly time-dependent once the goal is surpassed. Essay 2 examines how different types of entrepreneurial signals operate across the two stages of crowdfunding. In particular, it compares Spence-like signals, such as prototypes and prior entrepreneurial experience, with cheap talk signals, such as verbal and rhetorical communication. I estimate a sample-selection model via full-information maximum likelihood that explicitly distinguishes between the pre- and post-goal stages while accounting for the non-random nature of successful projects. Using a large sample of technology campaigns on Kickstarter, I find that Spence-like signals are especially important for achieving the funding goal, whereas cheap talk signals become more effective after the project has already met its target.