This paper develops a dynamic model of crowdfunding to characterize success rates and welfare and to identify optimal transparency and design policies. We also characterize average bidding profiles. Bidding costs generate two dynamic forces: (1) decreasing pivotality, driven by reduced scope for strategic complementarity as the deadline nears, pushes the slope downwards; (2) a news effect from observed bidding further pushes the slope downwards for concave cost distributions, but upwards for convex costs. These effects can explain prominent bidding patterns. Non-disclosure of funding progress yields higher welfare than full transparency given homogeneous costs. However, cost heterogeneity favours disclosure by enabling early bidders to activate otherwise passive, higher cost bidders. We also investigate the tradeoff between raising prices and thresholds and we demonstrate success and welfare gains from the indirect dynamic pricing permitted by current platforms.
