There are some limitations in this study. These limitations relate
to both methodology and data, which deserve more future research
efforts. For instance, to introduce the proxy for hotel online sales, we
assume that customers have a constant probability of posting a
review on a travel agent website. Though the validity of this proxy
was supported by empirical data, this assumption is still too strict
from a theoretical perspective. Future studies can be carried out to
investigate the possibility of loosing this assumption and re-test
the validity of the proxy. As well, although this study shows that
user-generated reviews have a significant impact on business
performance, we cannot be sure whether all reviews have the same
impact. Also, the authenticities of reviews are not considered either.
It would thus be useful to analyze the different influence or perceived
usefulness of different reviews. Furthermore, the effect of
the content of reviews on the performance of hotels merits investigation
using web-mining technologies to further understand the
impacts of online reviews. Findings of this study were based on the
data collected from a website in China. Therefore, the research findings
may not be generalized. Still, this suggests directions for further
research. Finally, the cross sectional analysis did not reveal time variance
characteristics, a panel data analysis in the future would be
beneficial to gain more insight into this topic.