Exploring the advertising elements of electronic word-of-mouth in social media: an example of game reviews

The influence of social communities has quietly surpassed traditional media. Electronic word-of-mouth (eWOM) is far greater than in other forms of traditional advertising. More and more enterprises hire key opinion leaders (KOLs) to write product-related comments, hoping to influence the purchasing behavior of other users in the community. In fact, the power of text reviews on social media is more powerful than traditional advertising models. For in-app advertising, it is one of the important issues to understand the focus of ad viewers to improve the effectiveness of advertising and then enhance the click-through rate (CTR) of in-App ads. However, relatively few studies focus on studying what elements should be contained in a successful commercial review on social media. Consequently, this study will treat social media reviews as a kind of new advertising modes and attempt to find the contained elements of ads in these text comments by using natural language processing (NLP), latent semantics analysis (LSA), and matrix diagram techniques. The discovered elements of positive comments (commercial reviews) will be compared to those in negative reviews (true authentic voices of customers). Based on the results, we can provide advertising companies with suggestions when hiring KOLs to write recommendation reviews.