Advertising and marketing in social media
Study on potentially problematic advertising and marketing practices in social media

On behalf of the European Commission, CentERdata, GfK and WUR (Wageningen University & Research) conducted research on advertising and marketing practices in social media. The purpose of this multi-method study was (1) to depict the social media provider landscape in Europe, (2) to assess advertising and marketing practices in these media with respect to European directives related to online consumer rights protection (Unfair Commercial Practices Directive), and (3) to identify possible remedies for any problematic practices identified.

Study methodology

Based on a large-scale preparatory study, which included (inter alia) online discussion groups with social media users and mystery shopping at social media’s advertising platforms, we identified (1) hidden advertising, and (2) the distortion of likes as common practices that are potentially problematic for consumers. To gain insight into how problematic these practices are and in the effectiveness of potential solutions, CentERdata developed four online experiments, which were run in six European countries among over 10.000 social media users.

Some findings

The results revealed, for example, that social media users generally have difficulty identifying “native advertising”, that is, advertising that resembles, in format and content, the non-advertising (organic) content that is published on the same platform. People who falsely identified a sponsored post as non-commercial content, rated that post as more credible and liked it better, as compared to people who correctly identified the same post as advertising. Consistent with existing literature, the inability to identify advertising as such seemed to prevent the activation of persuasion knowledge, which normally stimulates critically processing of the ad’s content. The label that is used in social media to indicate that a post is sponsored (e.g. “sponsored” on Facebook and “highlighted on Twitter in the Netherlands) appeared ineffective in all countries.

Figure 1. Examples of native advertising (Facebook)

The experiment also looked at the effectiveness of two interventions aimed at improving identification of native advertising as such:

  1. increasing the visual salience of the current disclosure label (e.g. “Sponsored”);

  2. highlighting the post to make it more distinct from regular content.

Highlighting sponsored posts did not improve identification, possibly because it was not clear users why the post was highlighted. A more salient label only appeared to improve identification of native advertising if the meaning of the label was unambiguous. That is, increasing label salience improved identification of native advertising in countries in which (a translation of) “advertised” or “sponsored” is being used, but not in countries in which (a translation of) the more ambiguous “highlighted” or “promoted” is being used. These results thus show that identification of hidden advertising can be improved by the uniform use of salient and unambiguous disclosure labels.

Figure 2. Effectiveness of interventions to improve identification of native advertising

The full report can be found here.

Related projects