On August 20, X user Ishan Sharma tweeted about his experience of tipping at a New York restaurant. “Tipping is such a scam in New York,” Sharma declared in a post. In a week, the tweet had 52 million views.
Compare this to tweets about some recent global developments. US politician Kamala Harris’ post announcing her presidential nomination has 3.6 million views since August 24. The Indian men cricket team captain Rohit Sharma’s post a day after winning the T20 World Cup on June 29 got 4.8 million views. Singing sensation Taylor Swift has 35 million views on a tweet posted on April 19 announcing her latest album.
Ishan Sharma is no celebrity. He is not running for the presidency nor does he have fans across the world. His LinkedIn profile says he is a 22-year-old engineering dropout who runs a YouTube channel where he “talks about career, freelancing and business”. It has more than 1.5 million subscribers. He has also co-founded a marketing agency that helps “startups and individuals grow organically on social [media] through various viral content marketing strategies”.
Why did a tweet about him having to tip more than he wanted to gain such huge traction? Experts who analyse trends on social media told Scroll that this is what “engagement farming” or “engagement baiting” on X looks like. As the name suggests, tweets that fall under this category, lure users into engaging with it. The term has gained more prominence since August last year, when X started its ad revenue sharing programme, under which verified users get paid a portion of the platform’s earnings from advertisements.
Kalim Ahmed, a researcher who studies misinformation, and the intersection of technology and culture, told Scroll that the intention of a user behind posting a tweet determines whether it qualifies as engagement bait. As an example, Ahmed cited other tweets that Sharma had posted during his visit to the US, to highlight a trend of making provocative posts.
“When he went to Harvard, he posted about the university’s merchandise being manufactured in Pakistan and about unisex toilets where one could get free tampons,” Ahmed said.
He claimed, “He is not interested in the debates such posts triggered, but in the engagement he receives as a result.”
X’s policy of paying users on the basis of engagement on their posts has encouraged engagement farming, experts said. “All social media users want traction on their posts, but now there is a monetary incentive involved with it,” said Prateek Waghre, executive director of the Internet Freedom Foundation.
Experts said that this, and other policy tweaks by X have marked a shift in the character of the platform. It now promotes viral content by encouraging users to tweet posts that are frivolous, provocative – or worse – deceitful and divisive.
The ad revenue sharing service on X is available to those who have verified accounts, more than 500 followers and at least five million cumulative impressions on their posts over the previous three months. Impressions are the number of times a post appears on users’ feeds. Since anyone can get a verified account on X by paying a subscription fee, getting paid from X is virtually a function of appearing on feeds of more users. This pushes users to post tweets that are more likely to go viral, experts said.
Joyojeet Pal, an associate professor at the School of Information at the University of Michigan, said that X has made a key change in how user timelines appear to boost content that has a higher propensity to go viral. In January last year, X rejigged how its two timeline tabs functioned. The earlier tabs “Home” (based on algorithms) and “Latest” (based on chronology) were changed to “For You” and “Following”. The “For You” tab is the default setting for X users.
“The concept of the ‘For You’ feed is directly lifted from [short videos platform] TikTok which curates content on the likelihood of virality, and not on the basis of how the algorithm views you as a person, what you would like to see and whom you follow,” Pal explained. “There is a clear pivot in the feed towards fluff, non-serious content posted by verified users whose tweets are boosted.”
Pal said that the influence of TikTok on X can be seen in how short, funny videos have become more frequent on user feeds.
But how much money could a user make from X, and how does the platform calculate the amount? Under its revenue sharing programme, the platform says that X gives users a portion of its ad revenues based on how many verified users viewed advertisements in the reply section of a post.
This system is both opaque and does not generate a substantial amount of money for users when compared to platforms like YouTube and Instagram, experts said. For starters, sharing revenue is a pinch for X as ad revenues for X itself have dwindled since billionaire Elon Musk took over the company in October 2022. In 2023, X’s ad revenues plunged nearly 50% from the previous year as advertisers turned away from the platform because of poor content moderation under Musk and the owner’s tweets that promoted anti-semitic and extremist views.
Moreover, there is no clarity on how the revenue sharing programme determines payments based on engagement on a post because the platform’s application programming interface, or API, is no longer available for free. The API allowed access to X’s user metrics, which was a useful tool for developers and researchers to study online behaviour. In February last year, Musk blocked free access to the tool.
“It is impossible to do an empirical study on which post generated how much revenue,” said Tavishi, programme officer at the Centre for Communication Governance at Delhi’s National Law University.
However, it is not without good reason that users are fishing for engagement on X even though earnings from the platform are not lucrative. Gaining prominence on X could open up possibilities for users to earn from other avenues, experts said.
Ria Chopra, a columnist who writes on pop culture and internet trends, told Scroll that even if engagement bait tweets do not generate a huge amount of money from X, they do add to a user’s social media following. “Once you have thousands of followers, you can use that to monetise in a lot of ways – you could run an ad campaign or start a podcast on another platform,” Chopra said.
Researcher Kalim Ahmed agreed. Citing the example of Ishan Sharma, he said it was likely that a lot of users who saw his tweet, went on to watch his YouTube videos to find out how a 22-year-old engineering dropout had made it big. “Some people also took out parts of his videos where he speaks of earning millions to call him out over not tipping, so he is getting engagement on YouTube due to what he posted on X.”
Pal of the Michigan University said it was also possible to make money by being “a proxy” for celebrities or politicians. “If I do not declare that I am in favour of a person and I am consistently putting up content that enables their perspective, then I can have a deal outside of Twitter,” he said.
The fact that ad revenues from X are contingent upon advertisements viewed in the reply section of posts is key in understanding the type of tweets that are intended to be engagement baits. Chopra noted that there has been an increase in tweets that are open-ended questions or pit one fan base against another. These tweets invariably end up in a war of words, which is precisely the motive, she said.
“Tweets that say things like ‘suggest three films for someone who knows nothing about Bollywood’ or those on the lines of why Delhi is better than Mumbai, have no intention other than getting engagement,” Chopra said.
Pal also said he had observed a rise in tweets that ask banal questions which tend to prompt easy, but divergent responses. A good example of this aspect is a tweet from May 2023 that got 1.5 billion views. The tweet read: “WITHOUT GOOGLING Name a famous historic battle”.
Pal said this trend marked the overall tone of discourse on Twitter shifting towards being more reactive. “You are either too happy, or too sad, or too angry, so you end up reacting in some way or the other,” he said. “Musk’s own tweets are proof of this.”
Experts pointed out this behaviour thrives in an uglier side of engagement baiting that operates on attracting angry responses. Ishan Sharma’s tweet could be seen as an example of this aspect called “rage baiting”, but it was harmless compared to posts that are intentionally racist or communal, said Ahmed.
The researcher gave the example of an X user by the name “Barry Stanton”, who gets millions of views on racist tweets, often targeted at Indians. The tweets portrayed Indians as dirty and abusive, and suggested that they belonged to race inferior to Caucasians. On August 29, the account was suspended, after it was reported by several users for violation of X guidelines. But the motive of the posts point to a trend.
“Attacking Indians is a popular way of getting engagement because there are millions of Indians on Twitter who get enraged and respond to such posts,” Ahmed said.
Another prominent example of “rage baiting” that played out on X recently was a photo posted on August 22 by a user flaunting her triceps with the caption “Brahmin genes”. The tweet got more than 7.5 million views. As several users weighed in both in support and opposition to the tweet, the caption became a trending hashtag. On August 25, an anonymous user got more than a million views on a tweet that drew a provocative comparison between “Hindu genes” and “Mughal genes”.
Kritika Goel, the India head of editorial operations at fact-checking website Logically Facts told Scroll that high engagement on provocative posts by far-right handles on X has been a noticeable trend. Posts targeting certain castes and religious communities are often a result of misinformation and disinformation being spread on X, she said. “During incidents like the Israel-Hamas war, the unrest in Bangladesh, the elections in India, we have seen X being a breeding ground for misinformation,” Goel added.
Evidence has emerged that at least in some cases, X has even incentivised misinformation. Goel cited a recent report by British non-profit Center for Countering Digital Hate which showed that X ran ads on five accounts that promoted misinformation after a mass stabbing in July in the English town of Southport. These five accounts got 260 million views in the week following the attack, the report said.
Even before the ad revenue sharing was launched for all Twitter users, Musk had made payouts to right-wing influencers as a pilot programme. More recently, a top executive at X said that the company’s move to hide “likes” on a post was aimed at promoting engagement on “edgy content”. In June, X also tweaked its rules to allow users to post adult and graphic content on the platform if they were consensually produced and labelled as such.
Tavishi, of the Centre for Communication Governance, said that the rise in provocative posts, hate speech and conspiracy theories on X since Musk’s takeover coupled with the monetisation of posts have resulted in a change in the character of the social media platform. “Twitter was conceived as a platform for political discourse as opposed to YouTube or Instagram which are more focused on content creation,” she said.
Pal of the Michigan University described this as “the single biggest issue in the social media ecosystem”.
He said there is currently no “equivalent of the old, serious Twitter” where users would discuss global and political matters. “What Twitter offered was a curation of news more than commentary,” he said. “You decided on what to read based on what showed up on Twitter. Now, you cannot rely on what shows up on Twitter.”