Providing a positive viewing experience is critical for content owners relying on advertising and subscription-based revenue models to capitalize on the opportunity present in online video. However, with more viewers turning to online sources for consumption, understanding the quality of experience (QoE) and providing an optimal QoE becomes more critical. In this paper, we will discuss a possible online video predictive model for improving QoE using global intelligence extracted from analyzing billions of streams in real time utilizing a big data processing platform. While predictive modeling has been used in conjunction with big data to analyze historical and current trends for countless other disciplines, its application in digital media delivery has not yet been explored. Our research shows that on-line predictive modeling can provide tremendous value to those looking to monetize and enhance the viewer experience.