The first money in

A casino-funded streamer hits the bonus round he has been chasing for six hours, and the number on his screen climbs past six figures. He screams; the chat erupts; he tips his chair back and grabs his head. By morning that fifteen-second moment is on a few hundred accounts you have never heard of — cut, captioned, sometimes slowed down for the gasp — running across TikTok, Reels, and Shorts. None of those clips is an ad, exactly. Each one is a lure: a glimpse of a life-changing win, with a soft current underneath pulling you back to the table where it happened.

I open here because it is the part of the clipping economy people would rather not look at, and because it is the most honest place to start. In the flagship piece I argued that clipping — pay-per-view Content Rewards — is a genuinely new way to buy attention: a brand posts a price per view and an army of strangers delivers it. This piece follows the money that got there first. And the money that gets to a new ad medium first is almost never the money you would put in a deck. It is the money that has no other door. The disreputable always master a new channel before anyone respectable does, for the simple reason that they have to.

Casinos wrote the playbook

Start with the biggest spender, because it tells you the whole shape of the thing. Stake.com is a crypto casino — bets placed in Bitcoin and stablecoins, settled on-chain, available wherever the licensing is loosest. In 2024 it booked roughly $4.7 billion in gross gaming revenue, which puts it among the ten largest gambling groups on earth — ahead of DraftKings.¹ A company most readers have never typed into a browser is bigger than the betting brand that runs the loudest US TV ads. It got there with almost no conventional advertising. It got there on streamers.

The mechanism is brutally direct: pay a popular broadcaster to gamble your money, on camera, for hours, in front of an audience that watches them win and lose in real time. Stake reportedly paid the streamer Trainwreckstv a staggering $360 million over the sixteen months to October 2022 to play on stream — by his own account, a sponsorship in the hundreds of thousands of dollars per hour.¹ It signed Drake, in a deal reported at around $100 million a year (an estimate — treat it as the order of magnitude, not the line item).² The wins are the product. The audience is the funnel. The streamer is the billboard that bleeds.

Then the platform risk arrived, exactly the way it always does in this economy. In 2022, after a streamer named ItsSliker was caught having scammed friends and viewers out of hundreds of thousands of dollars to feed a gambling habit, a #twitchstopgambling revolt led by Pokimane and Mizkif forced Twitch's hand; in September it restricted streams of unlicensed slots and casino sites.³ The channel that built Stake had been switched off at the platform's whim. Stake's owners did not lobby for an exception. They built their own channel. About two months later, Kick launched — a Twitch competitor controlled by Easygo, the company behind Stake, run by the same founders, Ed Craven and Bijan Tehrani — with a creator-friendly 95/5 revenue split and, conveniently, none of Twitch's gambling restrictions.³ When the rented stage gets taken away, the deep-pocketed grey-market advertiser does not leave. It buys the theater.

Clipping is the next layer down from all of this — the distribution tier. The big wins get cut into short-form and pushed to the algorithmic feeds, where a fifteen-second jackpot needs no context to travel, and curiosity does the rest: back to the live stream, and from there to the table. The whole sector has a marketing budget to match the need. Crypto casinos took in on the order of $80 billion in 2024, by one industry estimate⁴ — a vast pool of money attached to a product that, in most of its markets, cannot run a clean Meta campaign or buy a Super Bowl slot. Stake is merely the apex. Roobet, Rollbit, Duelbits, Gamdom, Shuffle — the names rotate, but the playbook is one document, and clipping is now its last page.

Prediction markets run the same motion

Now watch the identical engine drive a product that could not look more different on the surface. I write about prediction markets constantly, and I believe in the category — a real-money market is one of the better forecasting instruments we have built. But belief in the instrument is no reason to be naive about the marketing, and the marketing is the same one that sells a pop single.

In the 2024 US election cycle, Polymarket paid meme and finance influencers to post its odds — the #PMPartner push, where seemingly organic accounts began quoting the market's presidential probabilities to millions of followers.⁵ That was the rehearsal. The 2025–26 version was bigger and a great deal messier. Per a POLITICO investigation published June 5, 2026, Polymarket's marketing chief Matthew Modabber moved roughly $2.5 million to more than 800 recipients between January 2025 and February 2026 — including at least $350,000 to about two dozen influencers, who produced some 490 posts on X. Most were never disclosed as paid, and the money was reportedly routed through a personal PayPal account.⁶ Two days later, NPR reported that this was not one company's habit but the category's: both Polymarket and Kalshi had paid influencers to reshare posts, and both run dedicated clipping channels on Whop — the same rail the casinos and the music labels use.⁷

To be fair to the category — and I mean this as someone who keeps writing about it — none of that makes a prediction market a scam. The product is sound; the odds it surfaces are often genuinely informative. The point is narrower and, I think, more damning precisely because it is narrow: the distribution motion a serious financial product reached for, when it needed cheap reach in a hurry, was indistinguishable from the one a crypto casino reached for. Same rail, same undisclosed posts, same engineered organic feel. When the legitimate end of a category borrows the grey market's growth playbook wholesale, the playbook is the story.

Why the grey edge gets there first

This is not a coincidence to be explained away; it is structural, and it generalizes. A business that can run a clean Meta ad and list a SKU on TikTok Shop has the easy doors open to it, and uses them. A business that can't — because its product is gambling, or an unlicensed financial contract, or anything a payment processor and an ad network would rather not touch — has the deepest possible need for reach that is cheap, fast, and, above all, deniable. Clipping supplies exactly that. A clip that looks organic, posted by an account that looks like a fan, carrying no #ad and no obvious sponsor, is the entire value proposition. The grey-market advertiser is not tolerating the "looks organic" quality as a side effect. That quality is the product it is buying.

Which is precisely where the trouble lives. The feature that makes clipping irresistible to the businesses with no other door — that it doesn't look like advertising — is the same feature that puts it on a collision course with the rules about advertising.

The disclosure seam

The relevant law is not exotic. The FTC's Endorsement Guides, updated in 2023, are blunt: a paid endorsement must be disclosed clearly, conspicuously, and up front, and the responsibility runs both ways — the influencer who posts it and the brand that paid for it are each on the hook.⁸ Now hold that against the thing we just described. A clip economy engineered to look organic is, definitionally, a clip economy engineered not to look like advertising. The very design goal — deniability, the absence of an "ad" signal — is, in the eyes of the rule, the design goal of evading disclosure. This is not a grey area that happens to brush against the guides. It is a business model whose core feature and the regulation's core requirement point in exactly opposite directions. That is what I mean by the seam: the place where the clipping economy is most exposed, because its best feature is the regulator's clearest target.

And the enforcement is already arriving, predictably, on the gambling side first — the grey-market edge, again leading. Stake was forced out of the UK in 2025 after a Gambling Commission review of its local licensing arrangement.⁹ The UK's Advertising Standards Authority has been pulling gambling ads judged to carry strong appeal to under-eighteens. Roughly six US states moved to ban sweepstakes-style online casinos in 2025. And X quietly removed gambling from what its Paid Partnerships policy will permit. None of this is the FTC knocking on a clipper's door yet — but it is the same logic walking up the same street. Where the grey money concentrates, the rules follow; the gambling end is simply furthest along.

What's hard — and what the money is telling you

Here is the honest tension, and I want to sit in it rather than resolve it too neatly. This — the casino lure, the undisclosed odds, the engineered fandom — is the part of the clipping economy that gives the whole thing a bad name, and it deserves the reputation. The right intellectual move is to separate the mechanism from its early adopters: pay-per-view distribution is a neutral primitive, the same way a credit-card rail is neutral, and it will end up carrying music, software, and political campaigns far more than it carries casinos. The mechanism is not guilty of who used it first.

But the harder, fairer point is the one I do not want to wave away: don't pretend the early adopters aren't telling you something. The first money in is a signal, not noise. When the heaviest, fastest, most sophisticated spenders on a new channel are the businesses that profit from gambling on attention — and from a feed engineered to keep you watching one more clip — that is information about what the channel is unusually good at. The concern that clipping is, at its core, a machine for gambling on human attention is not a moral panic. It is a reasonable reading of where the budgets went first.

This is the same pattern I have chased before from the other direction. In Event contracts are derivatives and It's not gambling — it's information I argued that entertainment and gambling liquidity is what funds and builds the rails that serious uses later run on — the speculative money lays the track. Clipping is that thesis itemized as a marketing line. The casino budgets and the prediction-market influencer spend are, quite literally, the subsidy that is paying to build out and prove the pay-per-view distribution layer. The serious uses get cheaper, better rails because the grey market paid to lay them first. That is not an apology for the grey market. It is the mechanism by which it matters.

The grey market is the indicator

So follow the first money in, and you get two things at once. You get an X-ray of the channel — what it is genuinely good at, which is delivering cheap, deniable, recruiting attention at speed. And you get a map of where the regulation will bite, drawn for you in advance by whichever businesses are already drawing fire. The grey market is the leading indicator on both counts. Where its budgets go, the mainstream follows — the way sports betting marched from back-alley to stadium naming rights in a decade, the way nearly every advertising medium in history was pioneered by someone the respectable players would not be seen with, and then, a few years on, became the medium the respectable players could not live without.

The channel will grow up; they always do. The household brands and the disclosed campaigns are already arriving, and in a few years the casino origins will read like a footnote, the way nobody remembers that the early web ran on things the polite internet would not name. The only open question — the one the disclosure seam keeps asking — is whether the clipping economy, on its way to respectability, manages to acquire the honesty it was so conspicuously born without. That is the part worth watching. Not the froth of the next viral jackpot, but whether the rail learns to say, out loud, when it is selling you something.