A User Study of Netflix Streaming

• Published in: Design, User Experience, and Usability: Design Discourse pp. 481 - 489
• Main Authors: Jackson, France, Amin, Rahul, Fu, Yunhui, Gilbert, Juan E., Martin, James
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing 2015
• Series: Lecture Notes in Computer Science
• Subjects: Home network; HTTP-based adaptive streaming; Internet video streaming; Perceived video quality; Simulation modeling; User-Experience assessment; Video performance assessment
• ISBN: 3319208853; 9783319208855
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-20886-2_45

Netflix and Hulu are examples of HTTP-based Adaptive Streaming (HAS). HAS is unique because it attempts to manage the user’s perceived quality by adapting video quality. Current HAS research fails to address whether adaptations actually make a difference? The main challenge in answering this is the lack of consideration for the end user’s perceived quality. The research community is converging on an accepted set of ‘component metrics’ for HAS. However, determining an objective Quality of Experience (QoE) estimate is an open issue. A between-subject user study of Netflix was conducted to shed light on the user’s perception of quality. We found that users prefer to receive lower video quality levels first with marginal improvements made over time. Currently, content providers switch between the highest and lowest level of quality. This paper seeks to explain a better method that led to higher user satisfaction based on Mean opinion score values (MOS).