Using User- and Marketer-Generated Content for Box Office Revenue Prediction: Differences Between Microblogging and Third-Party Platforms
Tingting Song, Jinghua Huang, Yong Tan, Yifan Yu
Information Systems Research · 2019 · Volume 30 · Issue 1
Abstract
In this research, we build a prediction model of movie box office revenue by empirically exploring its intricate relationships with user-generated content (UGC) as well as marketer-generated content (MGC) on a microblogging platform and UGC on a third-party platform. Our analyses are based on a panel vector autoregression (PVAR) model that is calibrated with a combination of data from Weibo (microblogging platform) and Douban! Movies (third party). Our empirical results show that microblogging UGC (MUGC) is a significant predictor of box office revenue and has stronger predictive power than UGC on Douban! Movies (DUGC). In addition, we find that the volume of enterprise microblogs (i.e., MGC) predicts box office revenue directly and also indirectly via MUGC, and MUGC thus exerts a partial mediating effect on the predictive relationship between the volume of enterprise microblogs and box office revenue. Finally, a prediction model of box office revenue using lagged box office revenue, MGC, MUGC, and DUGC is proposed, and its forecasting accuracy is found to outperform that of existing models. Managerial implications on utilizing social media for enterprises are provided. The e-companion is available at https://doi.org/10.1287/isre.2018.0797 .
Keywords
box office revenue, (enterprise) microblogging, social media, third-party platform, user-generated content, marketer-generated content, PVAR
Enable JavaScript for citation tools and additional paper interactions.