Spreading the Word in a Digital Era

YAO, Song | SEILER, Stephan | WANG, Wenbo

The exchange of information online has become increasingly public, and there are numerous platforms that facilitate this type of interaction. The second generation of the world wide web (known as web 2.0) allows users to collaborate and share information via social media, blogging, and web-based communities. Before this, word of mouth (WOM) used to be confined within a small circle of offline contacts, with limited visibility.

This new era of WOM via popular social media platforms such as Facebook and Twitter raises the question of how strongly brand related online WOM affects product sales in the real world. Researchers Stephan Seiler, Song Yao, and Wenbo Wang explain that “one obstacle to answering this question is that casual inference is particularly difficult in the realm of online WOM because firms are not directly in control of the amount of WOM”.

To further investigate the impact of online WOM, the aforementioned researchers conducted a natural experiment, which leveraged a temporary shutdown of Sina Weibo—China’s most popular microblogging outlet—to estimate the casual effect of online WOM on the product demand for TV show viewership in mainland China and Hong Kong.

Their first step was to distinguish two possible behavioural channels that could drive the effect to WOM on product demand. First, WOM could complement product consumption, thus increasing demand because it enhances the utility from watching the show. Second, WOM could affect demand in a similar way to traditional advertising, where people are informed or persuaded to watch a specific show.

By exploiting the differences in the timing of microblogging activity, the researchers found that microblogging activity after a show has aired was the primary diver of viewership. They also found that the amount of activity before the show did not impact viewership. “This pattern suggests the complementarity between TV viewing and the consumption of (post show) microblogging content increases TV ratings”, the authors explain.

The researchers also wanted to investigate which type of post-show activity most strongly affected ratings, and discovered that microblogging activity expressing sentiment had the strongest effect. Interestingly, both positive and negative sentiment affected ratings, suggesting an engaging and possibly controversial postshow debate as the key driver of TV show ratings.

The study found that rather than trying to increase WOM before the airing of a show—to remind or persuade consumers—it is actually more important to foster an active discussion after the show. Additionally, the magnitude of the estimated WOM elasticity was significantly lower than previous studies, and therefore their findings caution against overstating the impact of WOM.

Although their research presented some interesting findings, the authors explain that, “similar to evaluating the effectiveness of other types of marketing activity, measuring the effect of microblogging using field data is challenging, and the correlation between microblogging activity for a product and its demand does not necessarily imply causality”.

They view their finding as important “because of the recent increase in marketing spending on social media and a general belief among marketing practitioners that WOM can be a highly effective way to reach customers”.

WANG, Wenbo

Associate Professor