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.


  • Share on