A Dynamic Structural Model of User Learning on the Mobile Internet
32 Pages Posted: 8 Oct 2009 Last revised: 5 Sep 2014
Date Written: January 29, 2011
Consumer adoption and usage of mobile Internet-based content services has been growing steadily over the past few years in many countries around the world. In this paper, we develop and estimate a dynamic structural model of user behavior and learning with regard to content generation and usage activities in mobile multimedia environments. Users learn about two different categories of web content – (i) content from regular social networking and community (SNC) sites and (ii) content from mobile portal sites. They can choose to engage in the creation (uploading) and consumption (downloading) of multimedia content from these two categories of websites. In our context, users have two sources of learning about how well the content matches their preferences – (i) direct experience through their own behavior and (ii) indirect experience such as the behavior of their social network neighbors. To incorporate these issues, we develop a dynamic structural model and estimate it using a unique dataset of consumers’ mobile Internet-based content creation and usage behavior over a 3-month time period. Our estimates suggest that, on an average, the content downloaded from mobile portal sites has the highest level of match value with user preferences. In contrast, the content downloaded from SNC sites has the lowest level of match value. Learning based on direct experience is more accurate (has less variability) than learning based on indirect experience. Our policy simulations suggest that mobile operators can dynamically tailor their content and incentivize users to generate content according to user characteristics and their preferences. Potential implications for mobile operators and mobile advertisers are discussed.
Keywords: structural modeling, mobile media, mobile portals, Internet websites, uploading content, downloading content, dynamic programming, simulated maximum likelihood estimation
Suggested Citation: Suggested Citation