CNN ran a segment this month on the “clipping” economy, the practice of chopping up podcasts, livestreams, and interviews into short-form video and flooding TikTok, Reels, and Shorts with them for a fee. One clip farm operator interviewed claimed to have made $30 million over his career doing it. Another called clipping the future of advertising itself.

Both claims are probably true in the narrow sense. Clipping has minted real millionaires, and brands really are shifting budget toward it. But “the future of advertising” is doing a lot of work in that sentence, because the entire pitch rests on a number that almost nobody along the payment chain has any reason to verify: the view count.

The model, in one sentence

Clippers get paid per 1,000 views on the clips they post, typically somewhere between 30 cents and six dollars depending on the platform and niche. That’s it. Not per real person reached, not per genuine engagement, per view, a metric that bots have been faking at industrial scale since long before clipping existed. Nothing about the payment structure distinguishes a view from a real teenager scrolling TikTok and a view from a script running in a server farm in a country the clipper has never visited.

That’s not a hypothetical risk. It’s a risk the industry’s own operators openly admit to.

The people running this admit it’s a problem

Daniel Bitton, the 18-year-old who runs the clipping marketplace Content Rewards, which pays out roughly $40,000 a day to clippers, told Forbes directly that bot fraud is “the single biggest threat” to his business model. His exact framing was that clippers can buy fake views to inflate their payouts, and that on a platform paying out that kind of daily volume, the incentive to game the system doesn’t go away. It’s permanent.

That’s the founder of one of these platforms saying, on the record, precisely the dynamic I’m describing: the incentive to cheat isn’t a bug that shows up occasionally, it’s a structural feature of paying people by the view.

The evidence that this is already happening at scale isn’t subtle either. Stake, the offshore crypto casino, ran what researchers have called one of the most aggressive clip farming operations ever documented, paying aggregator accounts to watermark unrelated viral content (wedding videos, fight clips, old photos, nothing to do with gambling) with its logo. The network ran through a three-tier system of Discord servers, a Whop marketplace, and coordinating communities, and it only accepted accounts already pulling 50 million or more monthly views, meaning the entire operation was built on top of pre-existing, professionalized engagement farmers rather than organic fans. One account tracked in the network went from 64,000 to 645,000 followers within months of switching to Stake content.

You can also see the fraud problem in what the industry has built to fight it. Multiple clipping marketplaces now advertise multi-layer bot detection, API-verified view counts pulled directly from TikTok and YouTube rather than self-reported screenshots, and geographic filtering to block what one platform calls “known view-farm regions,” naming Pakistan, India, and Indonesia specifically. Nobody builds an entire verification layer, markets it as a competitive advantage, and geofences specific countries against a problem that doesn’t exist. The fraud-detection cottage industry sitting on top of clipping is itself the strongest evidence that your skepticism is warranted.

Your incentive chain holds up

The clipper is paid on views. Buying some doesn’t cost much relative to the payout, and gets caught only if someone bothers to check. The clip farm bills its client based on aggregate views across its clipper network. More views, inflated or not, means a bigger invoice. The client, usually an influencer or a brand running a campaign, is buying reach and relevance, and inflated numbers deliver exactly what they’re paying for on paper: their face everywhere, algorithmically amplified, looking undeniably popular. If the content carries sponsor placements, the sponsor is paying for exposure to eyeballs, and the influencer’s incentive to investigate whether those eyeballs are real runs directly opposite to their incentive to keep the sponsorship checks coming.

At no point in that chain does anyone get paid more for being honest about fraud than for looking the other way. The clip farm has the technical ability to catch it and, per Bitton’s own admission, treats it as an ongoing structural threat rather than a solved problem. The client has the least ability of anyone in the chain to independently audit it, and the weakest incentive to try. That’s just what happens when everyone in a chain gets paid on a number that only one link in the chain can verify, and that link is also the one collecting a percentage fee on the total.

Where “influencer fatigue” fits in

Audience fatigue is worth taking seriously alongside the fraud question, because they compound each other. If genuine audience interest in a given face is actually declining, inflated view counts become more valuable precisely because they’re the only thing masking that decline from the brands still paying for it. A real engagement drop-off and a fake engagement report can look identical to a client checking a dashboard, which removes the market pressure that would normally correct for a creator’s fading relevance. The fraud doesn’t just steal ad spend, it can actively hide the signal that would tell a smart buyer to walk away.

Why “the future of advertising” pitch depends on nobody checking

The economic case clipping makes to brands is brutally simple: traditional paid CPM runs $40 to $80 per thousand impressions, while clipping campaigns run as low as 30 cents to a few dollars. That gap is the entire sales pitch. But that arbitrage only works if the cheap views are worth something close to what expensive views are worth, and the honest answer is nobody outside the platform actually knows what fraction of any given campaign’s views are real, because the verification tools that exist were built by the same companies collecting a fee on volume.

This isn’t a new problem clipping invented. Programmatic display advertising has run this exact experiment for over a decade, and independent estimates have consistently put fraudulent impressions somewhere between 10 and 30 percent of total volume, with global ad fraud losses estimated in the tens of billions annually. Clipping isn’t a new frontier for advertising so much as the same fraud dynamic wearing a TikTok filter, arriving in an industry with even less regulatory infrastructure and far younger, less experienced operators running nine-figure budgets through Discord servers.

The honest counterpoint

The fairest pushback to my thesis is that not every buyer in this space is naive. Some of the more sophisticated clipping operations, according to people running them, have moved away from paying purely on view count and toward treating clips like real performance media, tracking which content actually drives clicks, signups, or purchases, and reallocating budget toward what converts rather than what merely racks up views. Bots don’t click, sign up, or buy anything, so a buyer disciplined enough to measure downstream conversion rather than raw reach has a real defense against exactly the fraud vector you’re describing.

That defense requires sophistication, patience, and a willingness to walk away from a vendor showing great top-line numbers, which is a lot to ask of an industry currently being sold on the premise that it’s cheap, fast, and already working. Most of the money flowing into clipping right now, judging by how it’s being pitched, is buying the view count, not the conversion. Until that changes, the flywheel you described doesn’t need anyone to be dishonest on purpose. It just needs everyone to keep doing what they’re already incentivized to do.