This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility: If you spend time watching online videos, you might feel like covering your ears. For some influencers, dropping f-bombs and other swear words is part of their brand, and their followers embrace it.
But new research from the McCombs School of Business at The University of Texas at Austin suggests most creators might think twice about using profanity. The reason: On average, it reduces user engagement. "To Swear or Not To Swear: Exploring the Role of Profanity in Influencer Marketing" is published in Journal of the Academy of Marketing Science.
For an influencer producing 100 videos a year, 14 off-color words per video could cost up to $79,000 in YouTube revenue, according to Vijay Mahajan, professor of marketing, and former McCombs doctoral student Mahdi Niknejad Moghadam, now teaching at The University of Texas at Tyler. But profanity's effects can vary by context, the study shows. It offers guidelines for when dirty words are especially damaging—and when, occasionally, they're not.
"There are no FCC regulations to monitor social media influencers right now," says Mahajan. "This research gives platforms, influencers and content sponsors a clearer picture of profanity's impact on user engagement." Despite widespread profanity among influencers, no research had examined its consequences among the viewers they're trying to influence, Mahajan says. "Profanity can be a cue of authenticity and community belonging, but it also goes against social and sometimes platform norms." He and Niknejad used machine learning to analyze 43,892 videos on a popular platform, shared by its 200 gaming influencers with the most subscribers.
The program looked for swear words included in a dictionary developed by three UT scholars, including psychology professor James Pennebaker. The program measured concentrations of profanity, counting the number of times words appeared in a single video and accounting for the video's length. It also gauged volume, video quality and vocal emotion expressed in each video.
"We used machine learning models to extract certain auditory characteristics, such as how loudly they speak," Niknejad says. "If I used the f-word and you just read the transcript of what I say, you might think I'm angry. But when you hear my actual tone, it's different.
Extract — continue reading at the source.
