One Key Platform Feature Helps Some Projects Win Crowdfunding

According to a model created by Cong and Xiao, having an all-or-nothing element allows fund-raisers to set higher prices—the amount people donate for what they get in return—with less risk of discouraging potential funders. This is because interested donors amass information as the campaign progresses, and high pledges are only collected for good projects that can reach the funding threshold.

The choice to rationally imitate the behavior of others without using one’s private information is known as an information cascade. It occurs when information inferred from others’ choices outweighs a person’s own knowledge.

Information cascades happen in many contexts. They can feed speculation in financial markets or cause YouTube videos to go viral. Suppose a diner decides to try a new restaurant based on positive online reviews. But when she arrives, the restaurant is empty and everyone seems to be eating at the place across the street. If she changes her plans and decides to follow the crowd, she is participating in an information cascade.

What makes behavior on crowdfunding platforms and the like different from that of a typical information cascade, Cong and Xiao find, is that on platforms with an all-or-nothing threshold, cascades happen in one direction only—in the choice to support the project. Consequently, financing novel projects and aggregating information can be efficient, they argue, challenging existing theories on information cascades.

An all-or-nothing system changes decision-making because people consider how their choices affect those who choose after them. In Cong and Xiao’s theoretical model, people who believe in a crowdfunding project always choose to support it because they know the final funding determination will reflect the wisdom of later decision-makers, minimizing their downside risk. If enough additional willing backers decide to fund a project, an information cascade may result, resolving the issue in the project’s favor.

Information cascades and all-or-nothing features also manifest themselves in VC rounds in which an entrepreneur approaches venture capitalists one by one, sequentially. But on crowdfunding platforms, the large investor base improves the efficiency of the process. And the theoretical insights should apply to other situations that also involve large numbers of people or “economic agents” making sequential decisions that have an economic effect. That happens in voting, for example—in national elections, early results announced in one state affect ongoing voting in another, and the election results in turn affect the economy. It also happens in fashion, when people wearing their new purchases in turn influence what others buy.

“Information production becomes more efficient, especially with a large crowd of agents, leading to more successes of good projects and weeding out some bad projects, and generally harnessing better the wisdom of the crowd,” the researchers write.