Modern entrepreneurship places strong emphasis on experimentation. However, while the learning benefits of this strategic approach are generally accepted, we have a limited understanding of its potential costs. Motivated by this gap, this dissertation investigates the role of experimentation in early-stage ventures. The first study develops a conceptual framework defining the construct and explaining the factors driving its impact on performance. I define experimentation as the disclosure of an incomplete product in order to obtain market feedback. I argue that learning, adaptation, and appropriability are the primary channels driving its effect. The second study builds a stylized model to illustrate the learning channel. Combining spatial differentiation and Bayesian learning, the model formalizes the notion that experimentation is optimal only when both market uncertainty and signal precision are sufficiently high. The third study analyzes the adaptation channel. Building on the idea that adaptation capability depends on organizational structure, I hypothesize that experimentation is evaluated positively only when it is combined with low formal structure. I test this argument on a proprietary dataset of venture pitches from a university-based venture competition. I analyze the factors driving judges' evaluations and find a negative interaction between experimentation and formal structure. The fourth study examines the appropriability channel. I hypothesize that, when formal intellectual property is weak, ventures face higher imitation risks and thus engage less in experimentation. I test this argument on a hand-collected dataset of US software ventures, exploiting the US Supreme Court decision Alice Corp vs CLS Bank International as a negative shock to patent protection. I find that, following the ruling, the affected ventures are less likely to experiment. This dissertation contributes to the literature on entrepreneurship, learning, structure, and appropriability, and offers implications for entrepreneurs, investors, and policy-makers.