Break the Internet with Emotion
What makes us more inclined to like and share certain pieces of online content, while others get lost and ignored within our newsfeeds? Each day online content is disseminated among millions of people from around the globe, yet it is very clear that some of this content achieves greater engagement than others. Take the classic 2007 YouTube video ‘Charlie Bit My Finger’ for example, which as of today has a staggering 880 million views on YouTube. What exactly is that key ingredient in the recipe for gaining such viral traction? Some may argue that the answer lies in emotion.
What Are Emotions and How They Are Measured?
In a previous blog post, ‘What Emotion are you marketing to?’, we touched on the nature of emotions and emotional theory. In summary, emotions can be defined broadly as physiological responses to things we experience or anticipate. These physiological responses implicitly inform appropriate reactions to different situations and experiences, resulting in improved decision-making from survival and social standpoints. Thus, emotions are used by our brains to better understand and navigate the world around us. Humans make sense of these physiological sensations through the construction of ‘feelings’, with which we subjectively classify and associate our emotional responses.
Emotions are primarily measured within two dimensions: valence and arousal. Valence is an evaluation of whether an emotion is negative (bad) or positive (good). Conversely, arousal measures the intensity of emotion, ranging from low (passive) to high (active). These emotional dimensions are illustrated below with numerous emotional states plotted on the axes.
Source: Chang, Yeh, Hsing & Wang (2019)
Emotional Engagement and Virality
So why does this matter? A seminal study published in 2012 by Jonah Berger and Katherine L. Milkman investigated how emotion impacted the virality of online content. In this study, thousands of New York Times articles were assessed using automated sentiment analysis and human coding to determine the emotionality and types of emotions expressed in each article relative to the number of shares (Burger & Milkman, 2012). The authors uncovered two key insights:
Online content that is positively valenced (elicits positive emotions) tends to be shared more often than negative-valence content. However, content that evokes either positive or negative emotions is generally shared more often than content that evokes no emotions at all.
Online content that evokes high emotional arousal (emotions with high intensity, such as excitement and delight) is generally shared more regularly than low-arousal content. This applies to both positively and negatively valenced content.
Berger recently obtained nearly identical findings in a subsequent study he did with Microsoft Researcher Daniel McDuff, which addressed the impact of emotion on virality in the context of video advertisements (McDuff & Burger, 2020). You can find a summarized version of the study here. Ultimately, the findings of these studies both indicate that arousal activation is the most important consideration for inducing dissemination of online content. High-arousal content is more likely to go viral than low-arousal content, regardless of valence. However, online content that provokes positive high-arousal emotions is optimal, as it was found to have had the highest association with virality overall (greater than content that provoked negative high-arousal emotions). Unsurprisingly, online content that elicited negative-valence, low-arousal emotions (such as boredom) was found to hinder the likelihood of sharing in both studies.
These findings align with the contents of the viral ‘Charlie Bit My Finger’ video that was mentioned earlier. Throughout the video, Charlie’s older brother clearly experiences strong, diametrically opposed emotions. Initially, Charlie’s brother appears to be happy while playfully engaging with his brother. However, once bitten hard by Charlie, the brother suddenly becomes visibly sad as he whines about the pain from the bite. This distress causes Charlie to laugh and, in turn, results in yet another emotional shift for the older brother, as he now joins Charlie in his contagious laughter. The prominent emotions expressed in this viral video are high in emotional arousal yet fall within both the positive (happiness) and negative (sadness) valence quadrants. Thus, this video seemingly supports the idea that high-arousal emotions (regardless of valence) have a high impact on content virality.
How You Can Use Emotions to Induce Dissemination
The aforementioned research has some valuable implications for marketers as it indicates how emotional engagement, and the type of elicited emotion, can be used to optimize the dissemination potential of brand-related content.
Ideally, marketers should strive to evoke positively valenced, high-arousal emotions such as excitement, inspiration and awe in their marketing content. Content that elicits emotions like these is associated with the highest rates of engagement and is, therefore, most likely to be shared and seen amongst the targeted audience. However, positively valenced emotions are not the only way forward. Certain negative emotions can also be utilized to promote content dissemination, but only when these negative emotions evoke high emotional arousal. Examples of these emotions include anger, anxiety and disgust. However, due to their negative nature, one should consider how to effectively apply them - thus, a socially responsible brand could evoke anger in its marketing content by exposing injustices that are relevant to the brand. A strong example of this is the recent ‘Save Ralph’ advertisement campaign that evokes strong negative emotions of shock and anger by depicting the harmful effects of animal testing on an animated, anthropomorphic bunny named Ralph. Keep in mind these negatively valenced emotions need to illicit high arousal activation in order to be successful tools in content creation that is destined to go viral.
Marketers can also benefit from these emotional effects in the visual elements of their marketing content by leveraging the power of emotional contagion. Emotional contagion describes the convergence of emotional states between individuals after either direct or indirect exposure. In a previous blog post, ‘How Smiling Faces Boost Sales’, we discussed the topic of emotional contagion and explained how the visual depiction of emotion in the form of a smile can impact a viewer’s emotional state and, ultimately, a brand’s sales. You can also evoke these desirable, high-arousal emotions that promote content dissemination (like excitement and anger) by presenting them in visual elements of brand content. An example of this would be depicting desired emotions on the faces of models in display ads. According to the contagion theory, marketers can converge the emotional states of their targeted viewers with these depicted emotions and resultantly increase the likelihood of these viewers spreading this content.
It is important to note that when appealing to consumer emotions in marketing content, one should always strive to make the brand an integral part of the content itself. By doing so, you can promote the sharing of a brand’s content while also boosting positive brand-related outcomes, such as improved brand evaluation and increased purchase. The concept of achieving both these effects simultaneously is described by authors Ezgi Akpinar and Jonah Berger (2017) as valuable virality. A good example of this is Coca Cola’s 2016 ‘Taste the Feeling Campaign’. In the video ad for the global campaign, Coke evokes strong positive emotions (mainly happiness and joy) by depicting people who are smiling, laughing and showing affection towards one another. However, in doing so, the brand also makes its products highly prominent throughout, thus depicting to the viewer how Coca Cola facilitates these positive experiences.
Measuring Emotion with Neuromarketing
Creating emotionally engaging content is one thing but how do you ensure that the content you have created is being received correctly? Using neuromarketing technologies to measure and assess the emotional engagement associated with your brand and its marketing content can be extremely beneficial. Using an Electroencephalography (EEG) machine, one can measure emotional responses to visual and auditory stimuli by recording electrical activity in the brain. In addition to this, you can also use facial coding software to measure the emotional reactions of consumers. Webcams are used to record an individual’s face and the software is used to analyse subtle facial changes (like slight movements of the eyebrows and mouth) that are reflective of certain emotional states. Finally, Galvanic Skin Response (GSR) devices are also commonly incorporated into neuromarketing studies and work by detecting changes in sweat gland activity via electrodes placed on the skin. These changes in sweat secretion are useful indicators that we can use to determine the degree of emotional intensity (arousal) that someone experiences in response to a stimulus.
It is evident that emotion plays an important role in content virality, and with Neural Sense you can understand whether or not you have managed to achieve the correct levels of valence and arousal in your content creation and just how your consumers engage with it.
If you would like to learn more about neuromarketing technologies and how Neural Sense applies them, visit our website. Alternatively, if you are interested in using our services to optimize your brand or marketing content, then please get in touch.
References
Chang, Y. Yeh, W. Hsing, Y. Wang, C. 2019. Emotions in the valence and arousal dimensions. PLOS One. Figure. https://doi.org/10.1371/journal.pone.0223317.g002
Berger, J. and Milkman, K., 2012. What Makes Online Content Viral?. Journal of Marketing Research, 49(2), pp.192-205.
McDuff, D. and Berger, J., 2020. Why Do Some Advertisements Get Shared More than Others?. Journal of Advertising Research, 60(4), pp.370-380.
Akpinar, E. and Berger, J., 2017. Valuable Virality. Journal of Marketing Research, 54(2), pp.318-330.
Comments