In viewing TV program or video content, a paradigm shift in viewer’s behavior is happening recently. An increasing number of viewers is using on-demand viewing free from time and place, and trust word of mouth or friend's recommendation rather than the TV promotion.
TV stations have to not only broadcast their programs, but also deliver the appropriate content to appropriate users in an appropriate manner. In this research, we focused on user's behavior in the services combining video-on-demand and TV review SNS. And two goals are set: (1)to reveal viewing behavior comparing to traditional TV viewing, and (2)to verify the effect of increase and expansion of user’s viewings by word of mouth.
To achieve these research goals, "user's behavior is different from traditional TV viewing," "user’s viewing triggers his/her review," and "user’s review triggers other user’s viewing" are set as hypotheses. Based on these hypotheses, we built the trial site "Teleda" and collected each subject’s behavior logs on the site to capture actual user's behavior. Moreover, questionnaires and interviews were conducted to capture the user's inner mindset or changes in behavior off the trial site.
From the trial results, the first hypothesis is verified because there is no close correlation between the number of views in the trial site and actual TV viewing rate. And another two hypotheses; user’s viewing triggers his/her review, user’s review triggers other user’s viewing are also confirmed.
In the user-based detailed analysis, it was found that there were two kinds of users, active and passive users in information publishing. And also found that there were two kinds of TV 4 programs in the program-based analysis. One tends to be catalysts of word of mouth or discussion such as documentaries. The other program is not like catalysts, such as variety show or drama.
From these findings, the loss of viewing opportunity is compensated by the word of mouth effect especially for the "heavy" program which is tended to be skipped in TV zapping.By making it easier for active users to reduce a barrier for information publishing, it is possible to create a circulation between viewings and viral communication. The user reputation system would fit in an effective information publishing.
For the future work, we are planning the next trial with a larger number of subjects in a longer-term to observe the formation of clusters and the social graph. In addition, we intend to introduce external SNS and user invitation function for more practical use of the Social TV.