Capitalising on the Affect of Negativity on Social Media on Individuals

This article was originally written as an assignment for New Media, and was nominated by Dr Trish Campbell for a Student Writer Award of Excellence. This article has been changed to fit the formatting of this site, but otherwise remains intact. The original article can be found here.

In 2014, The Atlantic, an American digital-news and magazine publisher, revealed to the general public that in 2012, Facebook had undertaken a research project, which directly manipulated the data that more than 600,000 users saw on their platform and monitored their response to the manipulated data and how they interacted with the platform in general ^[1]. Though there were questions regarding the ethics of Facebook’s study, and the journal responsible for publishing the findings to the academic community issued an apology for the methods used, the discoveries revealed by the researchers’ study have far-reaching implications for both users and creators trying to reach a target audience. Examining the effect of emotional contagions on users, the diffusion of ’emotionally negatively’ biased news, and emotional triggers utilised by traditional marketing strategies reveals the possibility of creating a unique and innovative marketing approach targetting users of new media social networking sites

Facebook’s Study

Facebook’s study, titled Experimental Evidence of Massive-scale Emotional Contagion Through Social Networks, ^[2] manipulated the data that 689,003 random users saw on their Facebook feed over a week. Out of these users, 310,000 had the number of posts they saw expressing positive or negative emotions reduced. Half of the 310,000 users (155,000) saw fewer positive posts, and half saw fewer negative posts. The remaining 379,003 users acted as a control group, which had posts proportionately removed, at random, with no consideration to expressed emotion^[2].

Though emotional contagion, defined as the “process of transferring one individual’s behaviors and emotions to another individual ^[4],” was known to exist by researchers, there were doubts about its ability to ripple across social networks, in person or otherwise. Moreso, until this experiment, there was no evidence that emotional contagions could spread through text only means. Facebook’s study revealed that both concepts were true, but also revealed that emotions did not have to be targeted to spread: an individual’s negativity, even generalised or directed to another, could affect “friends” whom they had never met in person before^[2].

Notably, while the data collected revealed that those who had negative posts filtered posted more positive posts, and vice versa, it also showed that negativity is more impactful than positivity. According to the chart to the right, those who had negativity reduced posted more negative words (~.05% less than control) than those with positive words reduced posted positive words (~.125% less than control). While those exposed to reduced positivity posted fewer negative words (~.02% more than control) compared to those who saw less negativity posted positive words (~.04% more than control), this implies that those exposed to reduced positivity withdrew from their social networks by either making fewer posts or posting more emotionally neutral content.

Though the change is relatively small, less than one per cent, the implications are vast. When one considers that in 2011 the “average internet user has 669 social ties,^[5]” which has only grown since then, and that study does not consider public posting in “public groups” or users interacting on multiple SNS platforms, a single emotional contagion has the potential to go viral, affecting thousands, if not millions, of people. This phenomenon is particularly true when a single word can act as a contagion to an entire online community.

Emotional contagions, online or otherwise, are not necessarily a bad thing. Organisations attempt to seize on the effects of positive contagions for “fostering team spirit, bringing about performance effectiveness, fostering transformational leadership, and creating desirable organizational cultures^[4].” However, given the unique ability of social media to spread emotional contagions rapidly and farther than ever before, New opportunities arise for businesses to profit off potential customers’ emotional states.

Bad News Travels Faster and Farther on Social Media

While the adage that “bad news travels fast” remains true, it drastically understates the speed and reach that bad news diffuses on social media. Although the concept of bad news travelling fast has been generally accepted, there has been little evidence to prove it until the rise of social media^[6]. In 2016 a pair of researchers, Anna Fang and Zina Ben-Miled, analysed thirty-eight high profile, emotionally biased news tweets for six days and learned that the tweets having a negative bias were more often shared (retweeted) than their positive counterparts.^[6] Furthermore, through their analysis, they determined that the average diffusion time for a news story on social media was three to four days. At this point, the story receives a drastic decline, garnering negligible amounts of shares and likes. “The total number of retweets… ranged from 24 to 10,360 [per tweet] while the total number of likes ranged from 14 to 29,233 [per tweet] ^[6].” This data shows that advertisers will have a higher profit to cost ratio from targeting negative news articles, or more widely, outlets with a bias towards producing negative news. Their advertisements will reach more people in the window of opportunity that each article is actively diffusing. The emergence of the new media phenomenon “doomscrolling” multiplies this effect.

Doomscrolling, defined as:

the practice of obsessively checking online news for updates, especially on social media feeds, with the expectation that the news will be bad, such that the feeling of dread from this negative expectation fuels a compulsion to continue looking for updates in a self-perpetuating cycle^[8]

is a new phenomenon with no research on the subject. Despite the lack of research, it is inferable from the data on emotional contagions and the diffusion of news that those who partake in doomscrolling may be particularly susceptible to targeted advertising. Especially, considering the role that emotion has been traditionally demonstrated to have on shopping choices.

Sad Consumers Purchase More

In 1982 Ronald E. Milliman performed a study in which he tested background music in a supermarket to test how various music tempos affected shoppers, and that discovered lower tempo music, [generally associated with relaxation or sadness] caused “gross sales [to increase] to $16,740.23 for the slow tempo music.” ^[10] While he also warned that “The results may not apply to all supermarkets, nor to any other market situation” ^[10], Milliman, perhaps inadvertently, lead the way in discovering how emotion affected sales. Further research by Knoferle et al. made the connection between mode and emotion (sadness is implied by song in minor modes, happiness by major modes) ^[11] and drastically improved upon Milliman’s initial research. While Milliman targeted one supermarket, for a short period of time, and considered only tempo, this new research, tested in three markets over four weeks, showed “only minor mode, slowtempo music favorably affected sales volume”^[11].

To further illustrate this point, A 2008 study showed that customers made to feel sad were willing to pay 3.7~ times more for a bottle of water than neutrally conditioned customers^[12]. While these studies focused on physical, brick-and-mortar, stores, they have staggering implications and present profitable opportunities when adapted for use with new media.

It is, however, essential to consider the results of a 2020 study, which notes that “that customers visiting the online store preferred to stay longer, exhibit approach behaviour, when fast tempo music was played in combination with a cool colour.”^[13] The same study also notes that “this study has strong business implication for virtual businesses that website atmosphere is used to slow down customers’ movements so that they will stay longer in- in-store, leading them to eventually browse more products and buy from the store”^[13] and acknowledges and supports Milliman and Knoferle et al.’s findings.

Convergence

By looking at the effects of negative emotions on social media, how they and news spreads, and traditional research on negative emotion’s affect on consumer’s purchases – unique strategies are developable to maximise profit, as well as ways to test some of the strategies, which utilise new media.

Strategy 1: Target Users, With Large Networks, and Their Networks While They are Experiencing a Negative Emotional Contagion

Acting under the assumptions that sad people buy more and sadness is contagious through social media, instead of merely targeting advertising on keyword searches (e.g. running shoes) it would be more profitable to also target advertisements towards those who are actively experiencing sadness. This optimisation is achievable by targeting users with negatively biased emotions in search terms, key terms in social media posts, and who click or visit trackable sites associated with sadness. However, it is essential to ensure that the product advertised to the targeted user is relevant to them. Products displayed should have been searched for in the last three to four days.

Furthermore, It should be possible to use the concept of emotional contagion spread to target the social networks of those who are likely to begin spreading a negative emotional contagion to their network. This can be achieved using the same principle as above but would allow for a wider dispersion of products to different users. However, since the spread of contagions have a minimal effect, excluding particular circumstances, and the goal would be breadth, not depth, this may be less effective.

An example of this in practice would be: Sally is posting about an upcoming funeral and is visiting sites, with embedded social media tracking, about managing grief. Sally has also been searching for formal dresses. By algorithmically harvesting personal information about Sally, such as size, and tastes, it is possible to advertise a dress suitable to her needs, which she would be more likely to buy and pay a higher price for. It would also be possible to target those who respond to her posts, who indicate they know and liked the deceased, and are saddened by hearing about their death.

Strategy 2: Instead of Targetting Generic News, Target News With a Particularly Negative Bias, and the Users who Spread it and Search for Those Topics

It is known that bad news spreads fast, and as such, news outlets focus a good portion of their efforts to generating “clickbait” news with a negative bias. Thus, site-wide, user-specific advertising on news sites is both normal and relatively expensive. However, it is possible to target the coverage of specific stories or topics. By following trending topics on social media and news sites, particularly those with a negative bias, it should be possible to take advantage of the diffusion cycle and generate more effective targeted ads, displaying and costing less but reaching more active consumers.

This can be increased by studying the habits and targeting doomscrollers, who actively share the stories and participate in the diffusion cycle. In part, this is because they are likely to view more content, making it easier to target them, and in part, by analysing their habits, it may be possible to predict upcoming trending topics.

Strategy 3: Embed “sad” Music From Artists Which the Users are Familliar With in Store Pages

Similar to how social media sites track what sites users visit through their “share” widgets’ functionality, Spotify allows advertisers, artists, and producers to see users’ musical tastes. This makes it possible for users to share their data through embedded widgets. Since it has become bad practice to have pages auto-play music, the user must be encouraged to begin playing music. However, playlists can be generated on the fly, based on users’ likes, with artists they enjoy, and because the widget displays the artist and song, if it is something the user is familiar with, there is a much higher chance that they will choose to play it.

By developing a database of low-tempo, minor mode, songs from tracks listed on direct streaming services – it would be possible to give users of a store page, a unique musical experience, which also creates an environment in which they are willing to buy more, stay on the site longer, and possibly pay higher prices for their products.

This strategy’s results could be further accentuated if a sites’ colour pallet was comprised solely of “cool” colours^[13] and let the users choose a pallet theme most appealing to them from a selection. Through simple A/B testing, on two fronts, this strategy is testable. First: show some users the music widget, and hide it from others. Second: of those who choose to use the widget, show higher prices to some and not to others.

Conclusion

Through analysis of emotional contagions on social media, the spread of news on social media, and the effect of music in traditional marketplaces, it was possible to come up with three marketing strategies which rely on and implement new media and modern innovations, which should be relatively simple for businesses to implement. With further research and testing of the principles behind the strategies, it should be possible to turn them into a standardised recipe for profit generation for sites that sell a wide range of products.

References

1.  Meyer, R. (2014, September 9). Everything we know about Facebook’s secret mood manipulation experiment. https://www.theatlantic.com/technology/archive/2014/06/everything-we-know-about-facebooks-secret-mood-manipulation-experiment/373648/.https://www.theatlantic.com/technology/archive/2014/06/everything-we-know-about-facebooks-secret-mood-manipulation-experiment/373648/
2.  Kramer Adam D. I., Guillory Jamie E., & Hancock Jeffrey T. (2014). Experimental evidence of massive-scale emotional contagion through social networks. Proceedings of the National Academy of Sciences of the United States of America, 111(24), 8788–8790. https://doi-org.ezproxy.ardc.talonline.ca/10.1073/pnas.1320040111
3. Anthony Quintano from Honolulu, HI, United States, CC BY 2.0 <https://creativecommons.org/licenses/by/2.0>, via Wikimedia Commons
4.  Banerjee, P., & Srivastava, M. (2019). A Review of Emotional Contagion: Research Propositions. Journal of Management Research (09725814), 19(4), 250–266.
5.  Hampton, K., Goulet, L. S., & Purcell, K. (2019, December 31). Part 3: Social networking site users have more friends and more close friends. Pew Research Center: Internet, Science & Tech.https://www.pewresearch.org/internet/2011/06/16/part-3-social-networking-site-users-have-more-friends-and-more-close-friends/.
6.  A. Fang and Z. Ben-Miled, “Does bad news spread faster?,” 2017 International Conference on Computing, Networking and Communications (ICNC), Silicon Valley, CA, USA, 2017, pp. 793-797, doi: 10.1109/ICCNC.2017.7876232.https://core.ac.uk/download/pdf/141923364.pdf.
7.  Drunks23. (2015). [Morbo announces the news] [Meme]. Reddit.https://www.reddit.com/r/futurama/comments/2e05cv/tonight_at_11/
8.  Dictionary.com. (n.d.). Doomscrolling.https://www.dictionary.com/browse/doomscrolling.
9.  YP_Studio. (n.d.). [Stressed lawyer business women working and writting in notebook on desk at office]. Shutterstock.https://www.shutterstock.com/image-photo/stressed-lawyer-business-women-working-writting-1349151809.
10.  Milliman, R. E. (1982). Using Background Music to Affect the Behavior of Supermarket Shoppers. Journal of Marketing, 46(3), 86–91. https://doi.org/10.1177/002224298204600313.
11.  Knoferle, K., Spangenberg, E., Herrmann, A., & Landwehr, J. (2012). It is all in the mix: The interactive effect of music tempo and mode on in-store sales. Marketing Letters, 23(1), 325–337. https://doi.org/10.1007/s11002-011-9156-z.
12.  Cryder Cynthia E., Lerner Jennifer S., Gross James J., & Dahl Ronald E. (2008). Misery Is Not Miserly: Sad and Self-Focused Individuals Spend More. Psychological Science, 19(6), 525–530.
13.  Anwar A, Waqas A, Zain HM, Daisy Mui Hung Kee. Impact of Music and Colour on Customers’ Emotional States: An Experimental Study of Online Store. Asian Journal of Business Research. 2020;10(1):104-125. doi:10.14707/ajbr.200077.https://www.magscholar.com/ajbr/ajbrv10n1/ajbr200077.pdf