Elsevier

Computers in Human Behavior

Volume 65, December 2016, Pages 285-294
Computers in Human Behavior

Full length article
Consumer responses to promoted tweets sent by brands and political parties

https://doi.org/10.1016/j.chb.2016.08.033Get rights and content

Highlights

  • Effects of social media advertising differ for brands and political parties.

  • Promoted tweets can have negative consequences for political parties.

  • An important underlying mechanism that explains SMA effects is persuasion knowledge.

Abstract

This study aims to understand how consumers respond to social media advertising (SMA) by focusing on promoted tweets sent by brands and political parties, and examines persuasion knowledge as underlying mechanism of these responses. Two online experiments with between-subjects designs, comparing the effects of SMA (promoted vs. non-promoted tweet) and the source of the tweet (political party vs. brand), were conducted. Study 1 showed that consumers rarely notice it when a tweet is promoted. Study 2 demonstrated that when a promoted tweet was sent by a political party, the recipient's recognition that the tweet was a form of advertisement (i.e., activated persuasion knowledge) reduced online behavioral intention, increased skepticism, and negatively affected source trustworthiness and attitudes. This effect was not present for brands. Although research has shown that social media can be an important platform to engage audiences, this study is the first to study the mechanisms underlying the effects of SMA, and whether there are any boundary conditions to these effects. These findings suggest that political parties should be cautious in their use of social media advertising as it can evoke negative responses.

Introduction

Social media are important in our daily lives, and have become an important platform to send across messages to specific audiences, such as consumers and voters. Research has shown that the usage of social media, such as Twitter, can be useful for political parties, organizations, and brands to engage audiences online (Kruikemeier, 2014, Okazaki et al., 2015, Park, 2013, Van Norel et al., 2014). Therefore, Twitter has become an important medium for social media advertising (SMA; Van Dijck, 2011). Specifically, Twitter has allowed advertising since 2010, by including sponsored content, such as ‘promoted tweets.’ Promoted tweets are tweets purchased by advertisers that are visible to a specific target audience (Twitter, 2014). The deployment of promoted tweets can effectively evoke engagement (i.e., mentions of the brand in the tweet) and tweets with a positive sentiment about the advertiser (Dacres, Haddadi, & Purver, 2014).

However, the introduction of promoted tweets also involves some risks. Consumers may not appreciate this type of advertising (Van Dijck, 2011). Research suggests that the use of promoted tweets could negatively affect consumers' attitudes towards the brand sending it and lowers consumers' intention to click on a URL in the tweet (Wood & Burkhalter, 2014). Therefore, it is crucial for both advertisers and Twitter itself to understand whether consumers accept advertising on Twitter (Taylor, Lewin, & Strutton, 2011). Moreover, little is known about the mechanisms underlying the effects of SMA, and whether there are any boundary conditions to these effects. Therefore, this study aims to test these, by gaining insights into how consumers process, recognize, and respond to promoted tweets, and whether this is contingent upon the source of the tweet (i.e., a political party or brand).

As SMA is fairly new, little is known about the extent to which consumers understand this type of advertising, and thus whether or not they have developed persuasion knowledge about SMA. Persuasion knowledge refers to personal knowledge and beliefs about advertising motives and tactics (Friestad & Wright, 1994). The activation of persuasion knowledge in response to advertising is often seen as an important underlying mechanism that may explain different consumer responses (Ham et al., 2015, Lee and Hansen, 2013). This means that the use of persuasion knowledge in response to promoted tweets may be the underlying mechanism to the persuasive outcomes of promoted tweets (i.e., online behavioral intention) and consumers' responses towards the promoted tweet and its sender (i.e., source attitudes, source trustworthiness, and skepticism). Therefore, this study investigates consumers' use of persuasion knowledge, operationalized as the extent to which consumers recognize promoted tweets as advertising.

A possible boundary condition for the effects of promoted tweets could be the source that is sending out the tweet. Although promoted tweets are usually employed by brands, they are also used in political campaigns. Consumers may respond differently to SMA sent by political parties compared to SMA sent by brands. Recently, researchers have argued that voters and consumers cannot be treated in a similar manner and that marketing strategies for these two audiences should be implemented in different ways (Van Steenburg, 2015). As the decision to support a party is fundamentally different from the decision to buy a product, political advertising might affect voter attitudes differently than product advertising. Such insights might help to understand whether a ‘voter as consumer’ paradigm exists (Van Steenburg, 2015). Thus, tweets from different sources, and about different categories of goods and services, may have diverse effects. This study examines whether or not consumers appreciate the use of SMA in both contexts, and thus whether this is a beneficial strategy for both brands and political parties.

Altogether, by conducting two experiments, we try to provide insights into: (1) consumers' responses to promoted tweets (vs. non-promoted tweets) and the role of persuasion knowledge (i.e., the recognition of advertising) in the processing of this advertising format, and (2) whether these responses differ between promoted tweets sent by brands and political parties.

Section snippets

The use of persuasion knowledge in response to promoted tweets

The Persuasion Knowledge Model (Friestad & Wright, 1994) explains how consumers respond to different advertising messages. The model suggests that the way consumers cope with advertising messages depends on the recognition of the persuasive nature of the message. The development of persuasion knowledge is influenced by how much experience a person has with particular persuasion attempts and continues developing throughout the life span but is presumed to be well-developed in adulthood (Friestad

Consumer responses to the tweet and its source

Recognizing the persuasive nature of a message can have several effects on the way the receiver responds to this message. Empirical research and theory on persuasion knowledge has indicated that the activation and use of persuasion knowledge influences its persuasive outcomes (e.g., Boerman et al., 2012, Campbell and Kirmani, 2000, Fransen and Fennis, 2014, Wojdynski and Evans, 2016). In this research, we examine the extent to which the activation of persuasion knowledge may be an important

Design, participants, and procedure

To test our hypotheses, we conducted an online experiment with a 2 (promoted vs. non-promoted tweet) x 2 (source: political party vs. brand) between-subjects design. Power analysis showed that a total sample size of at least 179 participants was required to reach a statistical power level of 0.80 for detecting medium effect sizes (effect size f = 0.25) with an alpha level of 0.05 (Cohen, 1988). In total, 229 college students (81% female; Mage = 21.31; 46% owned a Twitter account) completed the

Randomization study 1

The four experimental groups did not significantly differ with regards to product interest, F(3, 225) = 0.095, p = 0.963; familiarity with the source, χ2 (3) = 3.90, p = 0.272; having a Twitter account, χ2 (3) = 0.85, p = 0.839; the frequency of Twitter use, (F(3, 225) = 55, p = 0.652; gender, χ2 (3) = 0.4.74, p = 0.192; and age, F(3, 225) = 0.42, p = 0.742. Despite the random assignment, voting intention for the political party did differ marginally between the groups, F(3, 225) = 2.41, p

Conclusions study 1

Promoted tweets look nearly identical to regular tweets; the only difference is the presence of a label and icon that informs consumers that this is a paid advertisement. Study 1 showed that a large percentage of participants did not recall this informative label. Consequently, the use of persuasion knowledge did not differ between the promoted and non-promoted tweets. However, when comparing consumers who did notice the ‘Promoted by’ label to the consumers who were exposed to the non-promoted

Design, participants, and procedure study 2

The procedure and stimuli used in the second experiment (conducted in September and October 2014) were similar to the ones used in the first experiment, with several adjustments. For reasons of clarity and conciseness, we will only discuss the differences between the studies. The experimental design of Study 2 included two brands and two political parties, making it a 2 (promoted tweet vs. non-promoted tweet) × 4 (source: two political parties vs. two brands) between-subjects design. By

Randomization study 2

The eight experimental groups did not significantly differ with regards to voting intention for the two political parties (Groen Links, F(7, 258) = 1.03, p = 0.408; PvdD, F(7, 258) = 1.13, p = 0.408); product interest (car F(7, 258) = 1.70, p = 0.108; washing machine, F(7, 258) = 1.11, p = 0.354); familiarity with the source, χ2 (7) = 8.41, p = 0.298; having a Twitter account, χ2 (7) = 5.68, p = 0.578, age, F(7, 258) = 0.38, p = 0.914; and interest in the environment, F(7, 258) = 0.44, p

Conclusion study 2

Study 2 replicated the effects of a promoted tweet on persuasion knowledge, and showed that this effect only occurs when the tweet is sent by a political party. When a tweet is sent by a brand, consumers recognize it as advertising (i.e., activate persuasion knowledge), regardless of it being promoted or not. But, when a political party sends a tweet, consumers seem to be more likely to use their persuasion knowledge when it is promoted. In addition, Study 2 replicates the effect of persuasion

General conclusion and discussion

The purpose of this study was twofold. First, it examined consumers' responses to promoted tweets and the role of persuasion knowledge as a potential underlying mechanism to explain possible effects. Second, it examined boundary conditions for the effects of promoted tweets by testing whether responses to promoted tweets differ when they are sent by brands and political parties.

In Study 1, we found that consumers rarely noticed the ‘Promoted by’ label that denoted a tweet as sponsored. Some

Research was conducted at

Amsterdam School of Communication Research (ASCoR), University of Amsterdam.

Acknowledgments

This project is supported by the Personalised Communication project at the University of Amsterdam (http://personalised-communication.net/).

The authors would like to thank dr. Eva van Reijmersdal for her valuable suggestions to improve this paper.

References (52)

  • G.C.I. Bruner

    Marketing scales handbook. A compilation of multi-item measures for consumer behavior & advertising research

    (2009)
  • M.C. Campbell et al.

    Consumers' use of persuasion knowledge: The effects of accessibility and cognitive capacity on perceptions of an influence agent

    The Journal of Consumer Research

    (2000)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • S. Dacres et al.

    Characterizing brand advertising strategies on twitter [White paper]

    (2014)
  • A.J. Flanagin et al.

    Perceptions of internet information credibility

    Journalism and Mass Communication Quarterly

    (2000)
  • M.L. Fransen et al.

    Comparing the impact of explicit and implicit resistance induction strategies on message persuasiveness

    Journal of Communication

    (2014)
  • M. Friestad et al.

    The persuasion knowledge model: How people cope with persuasion attempts

    Journal of Consumer Research

    (1994)
  • M. Friestad et al.

    Everyday persuasion knowledge

    Psychology and Marketing

    (1999)
  • R. Gibson et al.

    Changing campaign communications: A party-centered theory of professionalized campaigning

    The Harvard International Journal of Press/Politics

    (2001)
  • C.D. Ham et al.

    How to measure persuasion knowledge

    International Journal of Advertising

    (2015)
  • A.F. Hayes

    Introduction to mediation, moderation, and conditional process analysis: A regression-based approach

    (2013)
  • G. Hervet et al.

    Is banner blindness genuine? Eye tracking internet text advertising

    Applied Cognitive Psychology

    (2011)
  • J. Hsieh et al.

    Exploring the disseminating behaviors of eWOM marketing: Persuasion in online video

    Electronic Commerce Research

    (2012)
  • D. John

    Consumer socialization of children: A retrospective look at twenty-five years of research

    Journal of Consumer Research

    (1999)
  • A. Kirmani et al.

    Vigilant against manipulation: The effect of regulatory focus on the use of persuasion knowledge

    Journal of Marketing Research

    (2007)
  • R.N. Landers et al.

    An inconvenient truth: Arbitrary distinctions between organizational, Mechanical Turk, and other convenience samples

    Industrial and Organizational Psychology

    (2015)
  • Cited by (56)

    • Use of microblogging platform for digital communication in politics

      2021, Journal of Business Research
      Citation Excerpt :

      Such features are important in the context of India, which is a multicultural and multilingual nation. Third, the reception and influence of the tweets of the political parties and politicians are examined (Bakliwal et al., 2013; Boerman & Kruikemeier, 2016; Jungherr, 2013). Fourth, the sentiments expressed through the tweets are investigated to identify and discuss the prominent keywords used during the elections (Bakliwal et al., 2013; Boerman & Kruikemeier, 2016).

    View all citing articles on Scopus
    View full text