Multi-Platform Media Consumption Research: A Study on World Cup Audience

• Activity Number: 604
• Type: Topic Contributed
• Date/Time: Thursday, August 4, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and Marketing
• Author(s): Pengyuan Wang*+ and Elea McDonnell Feit and Eric Bradlow and Peter Fader
• Companies: University of Pennsylvania and University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
• Keywords: Multiple Media Platforms ; Cross-Platform Consumption ; Hierarchical Bayesian Model ; Posterior Predictive Validation

Abstract

We investigate media consumption on multiple platforms around the 2010 FIFA World Cup, with the aim of understanding and projecting multi-platform consumption habits of the United States based audience across digital properties including websites, streaming video and mobile. We estimate a hierarchical Bayesian model for the incidence data of people's usage of the three platforms during the World Cup tournament. The model is a micro-level probabilistic description of individual-level customer behavior and adjusts for possible external effects including time, tournament and game schedule. The effect of the game is modeled as a function of latent teams' media attractiveness parameters, and allows for us to understand the interplay between each team's FIFA world-rankings and its media attractiveness. Through a correlation structure, the model also allows us to consider the relationship between the platforms. MCMC method is used to sample the parameters and posterior predictive validation is used to check estimations from the model with both reach and frequency and more sophisticated and stringent tests using other characteristics of the viewers' behavior.