Involvement Trajectories on a Social Media Platform

• Activity Number: 590
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and Marketing
• Author(s): Lynd Bacon*+ and Peter Lenk and Danielle Murray
• Companies: Loma Buena Associates and University of Michigan and Shutterfly, Inc.
• Keywords: social media ; evolution of user engagement ; latent trait model ; Hierarchical Bayes ; on line marketing ; ecommerce

Abstract

How engaged users are with web sites and on line services is of practical importance to the organizations that own these assets. Users who don't interact with a site or service are unlikely to be affected by marketing efforts however well targeted. They are less likely to make purchases, take advantage of paid premium services, to make donations, or to engage in other kinds of revenue-generating activities. They are also less likely enroll others in platform use. We develop a dynamic, repeated measures, Hierarchical Bayes latent trait model of user engagement to measure its evolution over time. We apply our model to user behavioral data from a social media web site using MCMC procedures, describe the relationship between engagement and purchases made by the users, and discuss implementation issues. Our work extends and generalizes what we presented at JSM in 2009.