The assassination attempt on Donald Trump appeared to be a turning point in the public discourse around his candidacy. The Democratic Party suspended its advertising campaigns opposed to it, and politicians of all political persuasions condemned the senseless assassination attempt. And according to new insights shared exclusively with Forbes, this surge has resonated on TikTok, which is one of the top public platforms for discussing the upcoming election.
In the week after the shooting, the number of posts about Trump on TikTok increased by about 800%. The prospects for those videos increased from one billion to 6 billion. Before the shooting, most posts about Trump were critical of him, but posts from the week after the shooting showed him in a much more positive light.
Then the thrust went down again. TikTok users focused their attention on President Biden’s defection and Vice President Kamala Harris’ façade, and their sentiment toward President Trump dropped sharply, but it was completely back to pre-shooting levels.
The data, provided through a TikTok analytics company called Zelf, covers one of the most dizzying weeks in American politics in recent memory, quantifying TikTok speech for the first time in a way that may complement other comparative political research. , such as surveys and groups. And it offers one of the first data-driven looks at Vice President Kamala Harris’ candidacy, a virtual phenomenon that has captured the American political zeitgeist in recent days.
Beyond express candidates, Zelf also follows political issues such as inflation, immigration and climate change. Research over the past 15 days shows that (like the American public) TikTok users prefer Biden’s stance on guns to Trump’s, but they prefer Trump’s stance on inflation. This shows that some topics, such as Project 2025 and (perhaps surprisingly) the war in Israel and Palestine, simply haven’t been top topics of discussion on TikTok lately.
What dominated those discussions, according to the data, was a verbal exchange about whether or not Joe Biden and Donald Trump were compatible for president. The compatibility speech haunted President Biden in the weeks leading up to his resignation: In his press release, Zelf noted: “The suitability obviously serves as an anchor around the president’s neck, overriding any positive content on everything from women’s rights to climate change,” the Supreme Court. The company said that 32% of all interactions with Biden-related content were for videos that portray the president as unfit to serve, and that the topic is almost as prominent as other negative sentiments around all problems you tracked combined.
Zelf
Zelf gains his knowledge by collecting the total number of videos on TikTok that touch on that topic, CEO Pepijn van Kesteren told Forbes. TikTok also provides Zelf with statistics about videos, such as the number of times they have been viewed, liked, and commented on.
Due to TikTok’s restrictions on knowledge collection, Zelf is limited in the number of TikTok videos he can access. To measure conversations about Harris or any other topic, he pulls videos based on a TikTok “relevance” metric, which is a mix of recency and engagement (like likes, comments and shares) a video has received. According to van Kesteren, the company can only extract 10% of the videos on a given topic, but since it is the one that extracts the maximum videos with the maximum engagement, which usually represents around 85% of the overall engagement on the topic , giving the company a rich, if imperfect, read on how conversations on the platform are changing.
It’s from those videos that Zelf calculates its Net Promoter Score (NPS), the general metric it uses to assess how positive or negative speech is on a given topic at any given time. Net Promoter Scores is a common tool in marketing and advertising, and businesses use them to control how their name or logo is perceived.
Zelf’s knowledge shows a stark contrast in the NPS generated for Joe Biden (-31), Donald Trump (-5) and Kamala Harris (49). As Zelf explains, those numbers mean that positive posts about Harris get about 49% more engagement. than negative posts about him, while negative posts about Joe Biden get about 31% more engagement than positive posts.
Zelf
The company also provides scores on other civic and political topics, opening the door to a powerful transparency tool for researchers and journalists who can understand how TikTok shapes discourse about, for example, Taiwan independence or the wars in Ukraine or Gaza. .
It’s easy to see how Zelf’s knowledge can influence crusade decision-making. But whatever impact this data has at the presidential level, it will likely be even more useful for applicants at the state and local level, where ad buying and virtual recruiting campaigns can specifically replace how a candidate, or a certain position, is a candidate. Given TikTok’s abundant strength and popularity among the American public, a reliable TikTok discourse following can simply serve as a complement to polls, focus groups, and other classic analytical crusades.
In recent years, social media companies have reduced transparency tools that make conversation monitoring data available to journalists, researchers and the public. In 2021, Meta announced that it would be shutting down Crowdtangle, a realtime discourse monitoring tool that the company had previously offered to election officials in all 50 states, to help them detect potential election interference, and that the Stanford Internet Observatory had used to detect foreign influence operations. That same year, Twitter removed its Ads Transparency Center, which allowed members of the public to view political and issue ads running on Twitter. (In a previous life, I held content policy positions at Facebook and Spotify.)
TikTok has also deliberately limited the data it makes available to the public. In 2024, TikTok got rid of a tool that had been used by scholars and journalists to track the popularity of hashtags on the site after critics used it to raise questions about the platform’s influence. At the time, spokesman Alex Haurek said: “Unfortunately, some Americans and organizations have abused the Center’s studies to draw erroneous conclusions. Therefore, we modify certain features to ensure that they are used for their intended purposes. »
TikTok responded to a request for comment.