Anxiety and Ephemeral Social Media Use in Negative eWOM Creation


You have probably, at one time or another, found yourself in a situation where your favorite sports team just undeservedly lost a game and the sadness and anger you felt at that loss was overwhelming you. In the past, you might have vented those emotions in the living room, aiming your fiery rant at your friends or unsuspecting spouse. Nowadays, when consumers encounter “negative brand experiences” such as the loss of their favorite sports team, they respond with coping behavior (Duhacheck 2005) on social media, such as creating electronic word-of-mouth (eWOM) to complain (Stephens and Gwinner 1998) or to gain social support for their sentiments (Hennig-Thurau et al. 2004).


However, posting such eWOM causes stress and anxiety because individuals realize that their online content will be scrutinized by others (Krämer and Winter 2008). This may be reinforced when the message is negative, as this conflicts with an individual’s online impression management goals, namely achieving positive self-presentation (Berger 2014). Wakefield and Wakefield (2018) study this phenomenon in combination with the use of ephemeral social media. They apply the S-O-R theory (Mehrabian and Russel 1974), which states that stimuli (S) may alter consumers’ organismic state (O) leading to a response (R). In their research set-up,  brand experience is defined as the stimulus which alters consumers’ disposition in the form of message negativity and task anxiety, leading to a certain response depicted as message availability.

Blog post 2 1
Fig. 1 The S-O-R- theory as applied by Wakefield & Wakefield (2018)

It could be argued that the conflict between expressing negative brand experiences and self-presentation would lead to consumers attempting to avoid consequences of negative messages by reframing them more positively, using humor or staying silent. However, individuals facing an emotional experience appear to need to share it (Pennebaker, Zech, and Rimé 2001) even under social constraints (Pennebaker 1993). Therefore, the authors posit:

H1: Negative brand experiences will result in greater anxiety when creating eWOM compared to positive brand experiences

This leads to coping behavior in which the individual attempts to minimize the anxiety and stress caused by their negative expression, which is where ephemeral social media come in. Ephemerality facilitates coping with the conflict related to managing impressions through the limiting of message availability. This can, for instance, offer the perception of a smaller audience size (Ellison et al. 2011) or eliminate concern for the content’s future exposure. Based on this, the authors further hypothesize:

H2: The greater the negativity in a message, the shorter the message availability

H3: The greater the task anxiety, the shorter the message availability

The first study is conducted on a sample of 164 U.S. participants recruited through Amazon’s Mechanical Turk service. This sample is asked to create eWOM in a sports team setting similar to the one described in the introduction. The results provide preliminary support for the three hypotheses.

Blog post 2 2
Fig. 2 Relationship results from Study 1

However, in order to ensure generalizability, the experiment is replicated in a second study containing six other brand categories. Furthermore, post-task anxiety is now measured to see if setting a time limitation on negative eWOM will decrease anxiety. Additionally, the authors attempt to provide support for the impression management goals as a source of conflict, assuming that the greater the message availability, the more positive words for self-presentation purposes are used. Finally, the difference between objective economic brand experiences and subjective non-economic brand experiences is explored by having three search goods (clothing, automobiles, and restaurants) and three experience goods (banks, insurance and hotels) as brand categories. This leads to the addition of four new hypotheses to the first three:

H4. Post-task anxiety will be less than task anxiety         

H5. The greater the message availability, the greater the self-presentation.

H6a. Message availability for economic brand experiences will be greater than non-economic brand experiences as negativity increases.

H6b. Message availability for economic brand experiences will be greater than non-economic brand experiences as anxiety increases.


Blog post 2 3
Fig. 3 Task & post-task anxiety by message availability

The results from the second study replicate the findings of the first study with respect to H1-H3, thus demonstrating the generalizability of these relationships. Furthermore H4 is supported as anxiety decreased after restricting message availability, the effect being strongest for those who felt significantly greater anxiety during eWOM creation (fig. 3).



Blog post 2 4
Fig. 4 Message availability by message negativity

The same goes for H5, lending credence to impression management as an underlying concern. H6a is rejected, but H6b supported, indicating that consumers do differ in their restriction of message availability according to experience type (economic vs non-economic) with respect to the message negativity, but not with respect to task anxiety (fig.4).


Managerial implications

This study provides interesting implications for brand marketers interested in why and how eWOM can be constrained. To start with webcare management, this study suggests that a reactive strategy to negative eWOM may not be the optimal one. Although other research such as that of Proserpio and Zervas (2017) posits that management responses to negative feedback leads to less negative feedback, in the light of this study, management response would extend the availability of the message and prolong the incident for the consumer, discouraging the consumer to share his/her brand experience in the future. A better alternative is to address the anxious and negative consumers privately and/or offline, wherein ideally the consumer decides to share the resolution online of their own accord.

Furthermore, brand managers should realize that dependent on the type of good they work with, they should interpret online sentiment differently. For instance, brand managers of non-economic goods should be aware that negative online sentiment with respect to their goods is available for a shorter time and diluted with positive words and that consumers’ experiences could thus well be much worse than stated online.

Most importantly, the trade-off between impression management and negative eWOM creation is removed, or at least mitigated, when the consumer controls message availability. This means that a firm can solicit negative feedback through ephemeral social media in a win-win scenario where the company gets to improve their service/product whilst suffering the least possible amount of negative publicity and reducing their customer’s anxiety levels.


Strength & Weaknesses

A notable weakness of the article lies in the sample respondents, which were U.S. participants (n = 164) recruited through Amazon’s Mechanical Turk. Crowdsourcing respondents through MT has become a popular and easy way to find respondents, however there are several limitations to this method. Firstly, character misrepresentation may occur, as respondents have an incentive to falsify identity, ownership and activity information in order to qualify for a study (Wessling et al. 2017). Secondly, crowd-sourced respondents cannot be scrutinized and could thus be multi-tasking, or be interrupted, during the study. Furthermore, they self-select into studies and can quit at any time, thus MT workers may not provide reliable data or be particularly representative of the real-world consumers. Finally, the validity of data obtained from participants who have accumulated experience with social studies is questionable (Goodman & Paolacci 2017).

However, it must be noted that these are recent findings and that 43% of consumer research studies use MT workers as respondents (Goodman & Paolacci 2017). Moreover, the use of college students as study respondents has been critiqued for years before that, and these still seem to be seen as a viable respondent source.

The main strength of this article lies in its exploration of new drivers of eWOM characteristics, namely anxiety during creation and the enablement of limitations to message availability.  Furthermore, the authors immediately ascertained generalizability by replicating their first study on a broader array of brand categories, allowing for strong managerial implications to be made.



Duhachek, Adam (2005), “Coping: A Multidimensional, Hierarchical Framework of Responses to Stressful Consumption Episodes,” Journal of Consumer Research, 32, 1, 41–53

Ellison, Nicole B., Rebecca Heino, and Jennifer Gibbs (2006), “Managing Impressions Online: Self-presentation Processes in the Online Dating Environment,” Journal of Computer-Mediated Communication, 11, 2, 415–41.

Emmanuelle Zech, and Bernard Rimé (2001), “Disclosing and Sharing Emotion: Psychological, Social, and Health Consequences,” Handbook of Bereavement Research: Consequences, Coping, and Care, p.517–43

Goodman, J.K., and Paolacci, G., 2017. “Crowdsourcing Consumer Research”, Journal of Consumer Research, Volume 44(1) p.196-210

Hennig-Thurau, Thorsten, Kevin P. Gwinner, Gianfranco Walsh, and Dwayne D. Gremler (2004), “Electronic Word-of-Mouth via Consumer-Opinion Platforms: What Motivates Consumers to Articulate Themselves on the Internet?” Journal of Interactive Marketing, 18, 1, 38–52.

King, R.A., Racherla, P. and Bush, V.D., 2014. “What we know and don’t know about online word-of-mouth: A review of synthesis of the literature. Journal of Interactive Marketing, Volume 28(3), p.167-183

Krämer, Nicole C. and Stephan Winter (2008), “Impression Management 2.0: The Relationship of Self-esteem, Extraversion, Self-efficacy, and Selfpresentation within Social Networking Sites,” Journal of Media Psychology, 20, 3, 106–16.

Mehrabian, Albert and James A. Russell (1974), An Approach to Environmental Psychology. Cambridge, MA: MIT Press.

Pennebaker, James W. 1993, “Mechanisms of Social Constraint,” in Handbook of Mental Control. D.M. Wegner, J.W. Pennebaker, editors. Englewood Cliffs, NJ: Prentice Hall. p. 200–19

Proserpio, D. and Zervas, G., 2017. Online reputation management: Estimating the impact of management responses on consumer reviews. Marketing Science, 36(5), p.645-665.

Stephens, Nancy and Kevin P. Gwinner (1998), “Why Don’t Some People Complain? A Cognitive-Emotive Process Model of Consumer Complaint Behavior,” Journal of the Academy of Marketing Science, 26, 3, 172–89.

Wessling, K.S., Huber, J., and Netzer, O., 2017. “ MTurk Character Misrepresentation: Assessment and Solutions” Journal of Consumer Research, Volume 44(1), p. 211-230

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s