2021/05/16 by Bárbara Gomes Ribeiro, Ribeiro, Bárbara Gomes, Manoel Horta Ribeiro +7
Computer Science · Social Sciences · #Computers and Society (cs.CY) #FOS: Computer and information sciences #Hate Speech and Cyberbullying Detection #Misinformation and Its Impacts #Social Media and Politics #cs.CY
paper · pdf · doi:10.48550/arxiv.2105.07523
openalex publication_date 2021/05/16 · arxiv created 2022/02/25 · arxiv updated 2022/03/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In 2020, amidst the COVID pandemic and a polarized political climate, the Sleeping Giants online activist movement gained traction in Brazil. Its rationale was simple: to curb the spread of misinformation by harming the advertising revenue of sources that produce this type of content. Like its international counterparts, Sleeping Giants Brasil (SGB) campaigned against media outlets using Twitter to ask companies to remove ads from the targeted outlets. This work presents a thorough quantitative characterization of this activism model, analyzing the three campaigns carried out by SGB between May and September 2020. To do so, we use digital traces from both Twitter and Google Trends, toxicity and sentiment classifiers trained for the Portuguese language, and an annotated corpus of SGB's tweets. Our key findings were threefold. First, we found that SGB's requests to companies were largely successful (with 83.85% of all 192 targeted companies responding positively) and that user pressure was correlated to the speed of companies' responses. Second, there were no significant changes in the online attention and the user engagement going towards the targeted media outlets in the six months that followed SGB's campaign (as measured by Google Trends and Twitter engagement). Third, we observed that user interactions with companies changed only transiently, even if the companies did not respond to SGB's request. Overall, our results paint a nuanced portrait of internet activism. On the one hand, they suggest that SGB was successful in getting companies to boycott specific media outlets, which may have harmed their advertisement revenue stream. On the other hand, they also suggest that the activist movement did not impact the online attention these media outlets received nor the online image of companies that did not respond positively to their requests.