Description
Existing research of crowdfunding mainly focuses on the basic properties of the project such as category and goal, the information content of the project, however, is barely studied. By introducing Elaboration Likelihood Model into crowdfunding context and using a large dataset obtained from Kickstarter, a popular crowdfunding platform, we study the influence of project descriptions in terms of argument quality and source credibility, and investigate their impacts on funding success. We find information disclosed in project descriptions is associated with funding success. We also examine the practical impacts of project description by using a predictive model. Results show that our model can predict with an accuracy rate of 73% (71% in F-measure), which represents an improvement of 15 percentage points over the baseline model and 4 percentage points over the mainstream model. Overall, our results provide insights to researchers, project owners and backers to better study and use crowdfunding platforms.
Recommended Citation
Du, Qianzhou; Fan, Weiguo; Qiao, Zhilei; Wang, Gang; Zhang, Xuan; and Zhou, Mi, "Money Talks: A Predictive Model on Crowdfunding Success Using Project Description" (2015). AMCIS 2015 Proceedings. 37.
https://aisel.aisnet.org/amcis2015/BizAnalytics/GeneralPresentations/37
Money Talks: A Predictive Model on Crowdfunding Success Using Project Description
Existing research of crowdfunding mainly focuses on the basic properties of the project such as category and goal, the information content of the project, however, is barely studied. By introducing Elaboration Likelihood Model into crowdfunding context and using a large dataset obtained from Kickstarter, a popular crowdfunding platform, we study the influence of project descriptions in terms of argument quality and source credibility, and investigate their impacts on funding success. We find information disclosed in project descriptions is associated with funding success. We also examine the practical impacts of project description by using a predictive model. Results show that our model can predict with an accuracy rate of 73% (71% in F-measure), which represents an improvement of 15 percentage points over the baseline model and 4 percentage points over the mainstream model. Overall, our results provide insights to researchers, project owners and backers to better study and use crowdfunding platforms.