A Parsimonious Predictive Model of Movie Performance

Mustafa Canbolat, Kyongsei Sohn, John Gardner

OAI: oai:igi-global.com:264182 • DOI: 10.4018/IJORIS.2020100103

In this paper, the authors develop a parsimonious model that offers early prediction of potential success of a movie. In order to achieve this, a broad look at the drivers of movie success is required. Supply chain members will be making decisions regarding movie popularity with regard to licensing contracts, forecasting toy purchases, cross-promotions, etc. at varying times before a movie is released. A simple forecasting approach using publicly available data could improve supply chain decision making. Prior literature suggested that the virtual movie stock market, HSX, was a good predictor. Using a small set of variables including view counts, likes, and dislikes did offer some predictive value. However, HSX produces a forecast that dominates prior models while using a single readily available public data. Further, the HSX-based prediction showed consistency and convergence across a significant breadth of time.

Business Process Management Business and Management Business Information Systems