Reading between the lines: empirical studies on the relation between user-generated content and marketing Defended on Friday, 13 December 2019
Many consumers regularly post their opinion on products or services on
online platforms. This so-called user-generated content (UGC) reflects consumer
sentiment and can therefore be used for various purposes, among which the
prediction of firm performance measures like sales, or return on investment. A
firm performance measure that has not been studied yet in relation to UGC is
stock price volatility, which can be interpreted as a proxy of risk. In the first study
we examine the link between UGC and stock returns by investigating the
presence of shock and volatility spillover effects.
One of the most popular UGC platforms is Twitter. Apart from consumers,
companies contribute to the platform as well. The design of the network allows
for a quick spread of content and this has sparked the attention of marketers, as it
provides a new opportunity for viral marketing. Companies can either post
content via their own account, or by buying advertising space. In the second
study we investigate the effectiveness of advertising through promoted tweets
and promoted trends.
An important feature of Twitter is that they highlight the current ‘trending
topics’ among their users. In the third study we investigate these trends more
closely by modeling the messages regarding trending topics as time series and
estimating the corresponding distribution. By performing sentiment analysis on
the trending topic messages we can match certain distributions to certain types of
iii
sentiment. Moreover, by investigating when Twitter labels a topic a trend we can
study the influence of the platform itself on the spread of the trend.
Since consumer sentiment plays such an important role in UGC research, we
investigate in the fourth study whether it would be possible to actively manipulate
the sentiment in UGC that is posted online by conducting a field experiment.
These four studies will contribute to our understanding of the potential
influence of (advertising on) UGC platforms on firm performance, and to our
existing knowledge of the spread of social media trends. Finally, apart from
providing insights we hope to provide some tools for marketing managers to
actively influence consumer sentiment expressed in UGC.
Keywords
user-generated content, volatility, advertising, promoted tweets, promoted trends, trending topics, diffusion models, sentiment mining.