The clippening | The Verge

Earlier this year, after a tumultuous period serving as the former second-in-command at the FBI, Dan Bongino went back to what he is perhaps known best for: video podcasting. After Bongino exited the role in January, he began promotion for the return of his podcast, The Dan Bongino Show. He bought out a billboard in Times Square in New York; he dropped teaser videos for his first new episode in months. Bongino also deployed a more experimental promotional tactic, aimed at getting portions of his show in front of a wider audience. For this, he used clippers.

Clippers are largely anonymous social media accounts whose sole purpose is to rack up views. The accounts take a piece of longform content — an hours-long livestream, for example, or a podcast — and pull out the most exciting, controversial, or shocking moments. Sometimes the accounts are dedicated to clipping, but companies will also recruit accounts with existing followers. Clippers can be based anywhere in the world (one tech founder who uses clippers has described some of them as “hungry Slovakian teenagers”) while targeting English-speaking audiences.

After clippers get the source material that a brand wants to promote, they cut it down and blast their version into the open web. Hundreds or even thousands of clipping accounts might be sharing similar videos, all in competition with one another. You have perhaps learned about a TV show moment, a celebrity podcast appearance, or a new band via clippers without even realizing it; it just looks like someone sharing something. Clippers do not need to be affiliated in any real way with the subjects they are clipping, and the clipped content does not need to be creative, transformative, or even interesting. It is the cartilage of the internet, the placeholders for the algorithm to suck in and spit out.

According to a campaign listing on the service Clipping.net, The Dan Bongino Show started a 31-day campaign beginning the day after his podcast returned in February. There were few requirements or guidelines, only that clippers should pull moments from his new podcast episode, and include #danbongino in the video caption. The campaign ran across TikTok, Instagram, and YouTube, and clippers would be paid $150 for every 100,000 views (with funds dispersed via PayPal). A Discord message about The Dan Bongino Show pegged the budget at $2,000. Bongino’s team did not respond to a request for comment.

“It’s just a necessary marketing play that if you’re not doing you’re behind,” Clipping founder Anthony Fujiwara told The Verge in a message. “Clipping lets you abuse the algorithms of other platforms to grow your product exponentially.” Fujiwara says 62,000 clippers use his platform, earning $3,000 a month on average. Most are based in the US.

“We verify using their audiences as a metric for who we want to be a clipper,” he says. “Indian views don’t help anyone.”

If you are someone who wants attention, social media is just just another form of gambling in the age of algorithmic recommendation feeds. Creators and influencers can optimize their content or tweak titles and thumbnails, but ultimately they are all just pulling a virtual slot machine arm, hoping it will dispense views, engagement, and resultant revenue. For well over a decade, content creators have worked to reverse engineer “the algorithm.” Deploying clippers allows companies to gamble on content at scale, without paying a network of contractors upfront: Why bet once, when you could bet 50 times? Clipping is nothing new, despite the recent discourse around who uses it and why, and whether paying random accounts to share content promoting something is deceptive or manufacturing fake fandom. The reality is that more and more, the social internet is filled with clips, paid and unpaid, that stand in for the full-length podcast, video, film, album, or piece of writing. As online content increasingly becomes abstracted from the original work, what purpose does making the full version even serve?

It’s not just podcasters who hire what is essentially a personal army of microtask workers. Clipping.net also lists campaigns for TV shows like RuPaul’s Drag Race ($175 per 100,000 views) and Michael Carbonara, a candidate running for congress in Florida. (The instructions for the campaign note dictate “Your clips must NOT have Michael saying anything Anti-Trump / Anti-White House,” and note that AI-generated clips are acceptable, though.) The brief doesn’t include any instructions for disclosing it is paid content; the Federal Election Commission requires that digital content include disclaimers. Carbonara’s campaign did not respond to requests for comment. World of Wonder, the production company for RuPaul’s Drag Race, declined to comment.

On Vyro, another clipping service launched by MrBeast, Perplexity launched a campaign in early April centered around Joe Rogan’s use of AI. (Perplexity is a sponsor of Rogan’s show.) Clippers were instructed to make content based on Rogan discussing AI with guests like Bradley Cooper and Johnny Knoxville, with the AI company specifically mentioned. The campaign ran across Instagram and TikTok, paying $1.20 per one thousand views on a video — and came with more requirements. Accounts were required to have more than 10,000 followers to submit clips, and all posts were to include #PoweredByPerplexity and #sponsored (many clipping campaigns have no disclosures that the content is paid). Reached via email, Perplexity distanced itself from the clipping company, with spokesperson Jesse Dwyer saying Perplexity “has no knowledge” of Vyro and “takes any unauthorized use of the Perplexity name or logo very seriously.” When asked to confirm Perplexity had not run or authorized clipping campaigns, Dwyer stopped responding to The Verge. Vyro directed me to Evangelist, a platform that connects brands with clip farms, but the company declined to comment. If clips really are the standard for marketing — a tool that everyone uses, that is at this point old news — why the secrecy?

One of the biggest beneficiaries of the clip economy is Clavicular (real name: Braden Peters), a 20-year-old streamer who has enjoyed an unprecedented come-up thanks to short videos of him going viral. He has received mainstream news coverage of his maniacal focus on his appearance, in a fringe subculture known as “looksmaxxing.” He has thrown around racist slurs; hit his face with a hammer, saying he was literally chiseling his bones; and sang along to Ye’s song “Heil Hitler” with other right wing and manosphere influencers. You probably have never watched one of Peters’ hours-long streams — but I bet you’ve seen clips of them.

According to figures posted by Peters, more than 1,600 clippers farmed out content of him between March and April, posting nearly 70,000 videos that accumulated more than 2 billion views. They clipped him on a fake date with another influencer, in nightclubs with a rotating cast of women, and apparently being consensually choked until he convulsed.

“Everyone hating but as predicted the clip went giga viral,” Peters wrote on X of the choking stunt.

The clip ecosystem has a way of elevating even unknown personalities, reaching millions of people who will potentially never see where the original material clips come from. Peters, for example, has around 337,000 followers on Kick, a fraction of the following others command on the platform. Relatively few people watch his streams live, which are, frankly, uneventful — except for the moments he’s manufactured to be clipped later.

You do not even need to hire people to clip your content: many will do it for free, or you can just recycle everything you produce as clips. TBPN, the three-hour-long podcast popular among a subset of the Silicon Valley tech industry, gets only a few thousand views on YouTube, but most people watch the show via disembodied segments — clips of each guest — on X. OpenAI recently acquired the show, which generated millions of dollars via flashy, maximalist ads on the livestreams. As part of the OpenAI deal, TBPN will “wind down” the advertising, according to The Wall Street Journal.

Clippers by and large add nothing of substance to the original work — no analysis, no response, not even filters and music and cuts that fan edits often have. It is the most boring kind of content, spliced purely for the algorithm. And when it comes to paid clippers, it is hard to separate the aesthetics from the often very young editors who churn out content. Often, the only edit made to videos is a solid border around a clip with a clickbait-y few lines of text: “Joe Rogan talking about AI sounding too real 😨” one video on Instagram reads. The video is a clip from Rogan’s interview with Knoxville and includes the tag #PoweredByPerplexity and #sponsored, as per the clipping campaign rules. Another Instagram account that last posted in 2017 when it was sharing calligraphy videos appears now to be part of a clip farm, sharing Call of Duty clips tagged as sponsored with captions like “This might be the smoothest sniping has ever been in a battle royale 👀.” The original clip that made Clavicular into a viral personality was nothing special, simply a short video with the Kick logo in it. The caption, filled with words that to the average person meant nothing, was the actual growth hack: “Clavicular ran into a frat leader at ASU and got brutally frame mogged by him👀😂”

Now that the existence of clipping has hit mainstream consciousness, the strategy is being touted as the future of building platforms and growing a business online. Clipping is undeniably effective at generating views, which in the era of shortform video largely means the number of times a post comes up on a user’s feed (as opposed to requiring a certain watch time, for example). But whether clipping actually builds a meaningful and resilient audience is, to me, still unproven. The Rogan and Knoxville clip on Instagram, for example, generated 272,000 views, but almost no engagement: just over 700 likes, 14 comments, and 10 reposts. Did Perplexity’s clipping campaign meaningfully affect its business, or Rogan’s viewership? Did it change anyone’s mind about AI? Or was it simply the connective tissue between one scroll and the next, a piece of media that popped up and was forgotten just as quickly?

TikTok’s ascendance during the pandemic made shortform video everyone’s problem (or solution, depending on how you look at it). In less than a decade’s time, we are in a second pivot to video phase that now stretches beyond news organizations. Podcasts have turned into video talk shows hosted by journalists, influencers, comedians, and nuns. Political strategy firms flood Instagram, TikTok, and X with clips of politicians at events saying something shocking or stupid, knowing the clip itself is the news. Even I, a features writer, participate in the clip-ification of our digital lives: I make shortform videos explaining my stories to TikTok audiences who will mostly not end up reading the original piece. Some of it is out of necessity — if you make something for public consumption, the devil’s bargain is you have to promote it. But overindexing on the clipped version means eventually, the full-length content is a means to an end. If clips really are the present and future of media and reach online, one begins to wonder what justifies making the unclipped, complete content in the first place.

Now that the clipping cat is fully out of the bag, companies offering the service will likely be busy; maybe some firm really will hire a “Chief Clipping Officer” (though I’d advise against it). It might be effective in pumping views for a while — but if we are to believe the platforms where this kind of recycled content lives, reused clips may have a short shelf life online. Meta has said it’s cracking down on “unoriginal” content that includes many of the hallmarks of clippers: “adding borders, inserting captions, and changing the reel’s speed” are specifically called out. Clipping companies are pulling in millions of dollars by condensing politicians, podcasters, indie rock bands, and Silicon Valley technocrats into bite-sized content — but even the clippers need something to clip from. If all that matters is going viral, the value of producing anything more complete begins to diminish, and so does the viewers’ incentive to engage with anything beyond the clips.

In April, a clip of Tucker Carlson’s podcast circulated online, in which the right wing media personality said he would be “tormented” for playing a role in Donald Trump getting elected. The excerpt was shared by TMZ, discussed on The View, and clipped by Headquarters, the social media account run by former Kamala Harris staffers. What was part of the episode but that didn’t hit the clip farms was Carlson saying Barack Obama “hated” white people, or brushing off criticism that Trump is racist because he was known to be romantically involved with “everyone.” Carlson’s repentance for supporting Trump is a perfectly executed soundbite, just the right size to satiate a viewer who can then scroll to the next thing. The clip becomes more urgent than the thing it existed to promote.