Table 5 Literature review overview including information on communication elements and journey stages that were covered

Lütjens et al.⁵⁰

*

*

Unspecified

Meta analysis covering 20 years and 88 empirical studies.

Informativeness and entertainment show the strongest positive impact on consumers’ brand attitude towards online touch points.

Eigenraam & Verlegh⁴⁹

*

*

Shoes, Groceries, Recipe deliveries

Lab experiments and field study among workers, students and Prolific participants, measured with ANOVA and t-tests.

Informative content is for both warm and competent brands effective when it comes to perceived authenticity.

Emmers-Sommer &Terán (2019)

*

*

Health screening

Lab experiment with students (N = 357) analyzed with correlation matrix.

Celebrities are perceived more credible than health experts. However, consumers do not encourage celebrities alone to be advocating for health issues.

Hsu²³

*

*

Vaccines

Lab experiment with eye-tracking and survey among graduate students (N = 96), analyzed with T-test and MannWhitney U test.

Both celebrities and experts can effectively increase information awareness. However, the expert endorser increases the likelihood of message recall, while the celebrity endorser does not.

Jebarajakirthy et al.²⁸

*

Food

Lab experiment among Australian OTG consumers (N = 136), analyzed with T-test and MANCOVA.

Self-efficacy does not have influence on healthy eating intentions. Using gain frames and experts stimulates individuals to choose health on the go foods.

Shen et al.²⁴

*

*

Unspecified

Meta-analysis including 25 studies.

Narratives influence persuasion and persuasive content is especially effective when used with audio, video or long content formats.

Chang et al.²⁶

*

*

*

Electronics, Food, Travel, Clothing

Lab experiments among students in China, using ANOVA.

Irrespective of the narrative person, consumers generally prefer warm image advertising on WeChat over competent image advertising. However, for competent imagery, third-person narration tends to yield better results.

Keller & Lehmann¹⁸

*

*

*

*

Health

Meta analysis sampling 22,500 participants.

Case information, other referencing, female communicators in general positively influence intentions of health behavior. For low involvement audiences, fearful gain frames and other referencing positively impact health behavior.

Cavanaugh, Bettman and Luce¹⁵

*

*

Charities

Lab experiments with students (N = 206), analyzed with ANOVA.

Campaigns aiming to benefit distant others benefit more from using the emotion: love compared to hope, pride or compassion.

Berger & Milkman¹⁶

*

*

New York Times articles, Food, Travel

Field study sentiment analysis using LIWC and randomized survey experiment.

Positive emotion laden content drives more virality than other emotion laden content. However, both positive and negative high-arousal messaging enhance virality.

Ort & Fahr⁴⁸

*

*

Vaccines

Lab experiment with students (N = 447), analyzed with ANOVA.

Threat did not directly influence an individual’s attitude in favor of vaccination. Efficacy plays a more important role than threat in adopting health related behaviors.

Krishen & Bui (2014)

*

*

Food

Quota convenience sample of non-students at retail outlets, analyzed with ANOVA and ANCOVA.

The effect on indulgence intentions for fear appeals is higher than for hope appeals in the setting of non-indulgent food ads.

Heimbach & Hinz⁵⁵

*

*

News articles

Field study using sentiment analysis with LIWC

Virality is strongly associated with positively and emotionally written content, as well as anger-inducing content.

Berger et al.¹⁷

*

*

News articles

Multimethod approach with field data (NLP of over 35,000 pieces) and controlled experiments among Turk participants (N = 278). Analyzed using regression.

Emotions significantly contribute to virality, influenced by the intensity with which discrete emotions provoke arousal and uncertainty. While language that expresses anxiety, hope, and excitement captures attention, sadness tends to have the opposite effect.

Carey & Sarma²⁰

*

*

Road safety

Lab experiments with 62 and 81 male students. Analyzed using one-way ANOVA.

Fear in combination with efficacy in messaging significantly lowers driving speed of young male drivers. However, fear alone did not make an impact.

Forbes et al.²¹

*

*

Animals

Lab experiments with Australian students ( ~ N = 24). Analyzed with T-test and ANOVA.

Fear-related stimuli capture attention by purportedly facilitating rapid processing with minimal stimulus information.

Kim et al.¹⁹

*

*

Restaurants

Lab experiment N = 473, analyzed using confirmatory factor analysis (CFA).

Threat-generated fear significantly triggers consumer behavior change amid COVID-19, resulting in commitment to hygienic behaviors, prioritization of local restaurants and engagement in conscious consumption.

Herhausen et al.⁵⁸

*

*

S&P 500 companies

Text mining while using LIWC classification. Followed by one-way ANOVA.

The use of high-arousal-emotion in negative eWOM of brand communities increases its virality.

Griskevicius & Kenrick³⁵

*

*

Various such as travel and automotive

Literature review

Self-protection is triggered by angry faces and threats which results in increased tendency to conform. Affiliation is triggered by being with friends or coworkers and results in seeking products to connect with others. Kin care is triggered by vulnerable kids and presence of family members which results in increased trust and giving without expectation.

Okuhara et al.³⁷

*

*

Vaccines

Lab experiment with 696 Japanese respondents and analyzed with one-way ANOVA.

Both disease avoidance and kin care influence attitude towards vaccination in Japanese vaccine hesitant groups.

Chen & Bell (2021)

*

*

Health

Meta-analysis of 16 studies.

First-person-narrative enhances the persuasiveness of a health message, however, it does not have a positive effect on behavioral intention or perceived effectiveness of the message.

Kuipers et al.⁹

*

*

*

*

*

*

Online Health Communities

Online field experiment (N = 18,426) analyzed with multiple regression models.

The influence of communication elements is not the same in both awareness and engagement phase. Fear and self-protection are found effective in awareness stages, but the combination of love with kin care and affiliation is more effective for engagement.