Does Offline TV Advertising Affect Online Chatter? Quasi-Experimental Analysis Using Synthetic Control

Abstract

This study analyzes the impact of offline television advertising on multiple metrics of online chatter or User-Generated Content (UGC). The context is a quasi-experiment in which a focal brand undertakes a massive advertising campaign for a short period of time. The authors estimate multiple dimensions of chatter (popularity, negativity, visibility and virality) from numerous raw metrics using the content and the hyperlink structure of consumer reviews and blogs. The authors use the method of Synthetic Control to construct a counterfactual (synthetic) brand as a convex combination of the rivals during the preadvertising period. The gap in the dimensions of chatter between the focal brand and the synthetic brand in the test versus advertising periods assesses the influence of advertising. Offline TV advertising causes a short but significant positive effect on online chatter. This effect is stronger on information-spread dimensions (visibility and virality) than on content-based dimensions (popularity and negativity). Importantly, advertising has a small short-term effect in decreasing negativity in online chatter.

Keywords: User-Generated Content, Online Chatter, Synthetic Control, TV Advertising, Difference-In-Difference, Quasi-Experiments, Matching, Virality, Offline Advertising