The media consumption moving online was supposed to enable people to access information more effectively, instead, to some extent, people were yet overwhelmed. Clickbait, defined as “(On the Internet) content whose main purpose is to attract attention and encourage visitors to click on a link to a particular web page” (in Oxford English Dictionary), prevails and restrains people retrieve their desirable information effectively. In this exploratory study, we proposed two intriguing research conjectures, i.e. how rhetorical features in the clickbait influence the visiting traffic to the publisher at different levels and attempted to understand the antecedents and consequences of the prevalence of clickbait. In collaboration with a leading digital media company, we longitudinally collected a massive dataset in 2017. To test the research conjectures, we designed and developed a series of text mining methods and applied econometric analysis for empirical validation. The findings revealed 1) the rhetorical characteristics (hyperbole, insinuation, and visual rhetoric) could entice individual to click the baited headlines, 2) there was a quadratic (inverted U-shaped) relationship between number of clickbait posted by publisher and its visit traffic, and 3) such non-linear relationship was moderated by publisher's age.