This study explores the role of online enga
gement, homophily and social influence in
explaining traffic and news consumption by social
network users at an external news website.
The authors jointly model visits and page views
for a panel of users who registered with the
news site using their Facebook accounts. In th
eir model, the authors account for homophily
using a latent space approach, and account fo
r endogeneity, heterogeneity, and unobservable
correlates. The results show
that measures of an indivi
dual’s activity on Facebook are
positively associated with that individual’s actions at the news site. In addition, knowing what
a user’s Facebook
friends
do at the content website provides insights into a focal user’s
behavior at that website, as visitors with friends
who visit external news sites are more likely to
visit the news website studied. In
addition, news consumption (not
just visits) also depends on
friend’s actions but such an impact varies w
ith the individual’s underl
ying browsing mode. We
highlight the importance of social influence
in news consumption and further show that
homophily bias in news consumption is similar
to prior research in other categories. Our study
also highlights that visitors’ past browsing patter
ns are important predictors of future content
consumption, although social network informa
tion significantly impr
oves prediction beyond
the effect of such more traditional behavioral metrics. Finally, we find that Managers can use
readily available data for bot
h prediction and targeting.