YouTube had not touched the entry requirements for its Partner Program since 2018. On Monday it doubled them. From next February a new channel will need 8,000 watch hours or 20 million Shorts views to earn a share of ad revenue, and what YouTube is offering smaller channels instead of AdSense is shopping bonuses, brand deals and trend incentives.

We also went back to the World Cup, three and a half weeks after the final, to work out whose bet on it actually held. ThoughtLeaders tracked 212,912 videos about the tournament from 28,258 channels, and the brand that came out on top never signed a FIFA deal. In Creator on the Rise, we meet the reptile channel whose audience just turned out two and a half times over for a video defending its own hobby.

Plus the community platform that turns up in more than two hundred YouTube descriptions a day without buying an ad, and a week in which Speed and Kai Cenat have been trying to beat Minecraft on hardcore since Friday. Let's get into it.

This week's deep dive

YouTube Doubles the Bar for New Creators

On Monday, YouTube announced the first overhaul of its Partner Program entry requirements since 2018, and it doubled both performance thresholds. A channel applying for a share of ad and Premium revenue currently needs 1,000 subscribers plus either 4,000 public watch hours in a year or 10 million Shorts views in 90 days. From February 1, 2027 it will need 1,000 subscribers plus either 8,000 watch hours or 20 million Shorts views. The subscriber minimum does not move, and the lower tier that unlocks fan funding and shopping stays where it is at 500 subscribers. Channels already in the program are grandfathered in, though they have to accept the updated terms by January 31.

A second rule bites sooner for Shorts channels: to keep earning ad revenue from short-form, a partner has to hold at least 10 million qualified Shorts views every rolling 90 days. Fall under that and the Shorts payouts stop, though the channel keeps earning on its long-form videos. Existing members also have a new activity floor to clear: 1,000 watch hours a year, or a million Shorts views, or two long-form videos plus five Shorts every 90 days.

YouTube's stated reason is that the platform has simply grown past the old numbers. VP Amjad Hanif pointed to 200 billion Shorts views a day and more than a billion hours watched on television sets daily, and argued that a higher watch-time threshold will "mean higher payouts" rather than paying someone "a few cents for that month." That is defensible on its own terms: a revenue share spread across a wide pool of tiny channels produces checks nobody can live on. The company also says it has paid creators $100 billion over the past four years. What it amounts to in practice, though, is that ad revenue is being repositioned as a reward for channels that have already made it, and everyone below that line is being pointed towards commerce, shopping bonuses and brand partnerships instead.

For advertisers the consequence is a change in who picks up the phone. A creator sitting on 12 million Shorts views a quarter has a business today and will have a materially worse one in February, and the only revenue line available to them that does not depend on a YouTube threshold is a brand deal. That is a larger, more motivated, better-priced supply of mid-tail inventory arriving over the next six months, and it lands in the part of the market that has always been hardest to buy at scale because these channels have no rate card and no representation. The channels worth talking to first are the ones between the two bars, because they are the ones for whom a sponsorship stops being extra money and starts being the whole model.

Zooming Out

The 2018 thresholds were set when YouTube was arguing that anybody could build a business on the platform. Doubling them is the company deciding it no longer needs to make that argument, because the supply of content arrives whether or not the bottom half of the program gets paid. The part to watch is what replaces AdSense for those channels. Shopping bonuses and trend incentives are discretionary, unpublished and revocable in a way that a revenue share is not, so a creator's income becomes something YouTube decides rather than something it calculates.

The story behind the numbers

From Our Data

Who Really Won the World Cup in Influence?

Spain beat Argentina in the final on July 19, which is long enough ago that we can now do the useful version of this analysis: not who made the most noise during the tournament, but whose bet on it is still paying. Between June 1 and yesterday, ThoughtLeaders tracked 212,912 videos about the World Cup from 28,258 channels, gathering 35.7 billion views between them. Videos carrying disclosed paid promotions did 4.2 billion of those views, about 12% of everything the tournament generated. And it is still going: 19,957 of those videos were published after the final and have done 2.2 billion views since.

The more interesting question is who those 28,258 channels actually were, because most of them were not football channels a month before the tournament started. Sorting the 150 biggest of them by what they do gives five distinct groups, and they are nowhere near equal.

World Cup views by channel type, 150 biggest World Cup channels

General creators who pivoted · 78 channels · e.g. Celine Dept — 12.7B

Broadcasters and rights-holders · 29 channels · e.g. FIFA and CazéTV — 7.3B

Clip and edit channels · 30 channels · e.g. Eesh Khera — 2.7B

Football media channels · 8 channels · e.g. Futcrunch — 0.4B

Brand-owned channels · 5 channels · e.g. Kia America — 0.3B

Each bar is the total World Cup views for one type of channel, across the 150 channels in ThoughtLeaders' data with the most of them. The group that won is the one that was not in the category in May. Celine Dept, the Belgian football-challenge creator, posted 67 World Cup videos and did 2.2 billion views on them, with a single Short asking who would win the thing doing 173.8 million on June 15. Nick DiGiovanni, a food creator, published 32 and did 804.1 million views, led by a World Cup cake at 91.2 million.

The other four groups each did something specific. Broadcasters put entire matches on YouTube: Brazil's CazéTV posted 788 videos for 1.04 billion views, and its full 90 minutes of Brazil against Japan on June 29 did 171.9 million on its own. Clip and edit channels turned single moments into Shorts, and the size of channel doing it is the surprise, with Eesh Khera converting 99,300 subscribers into 73.2 million World Cup views and one Short about Ronaldo and Messi at 41.3 million. Football media channels went for volume and reliability rather than spikes; Futcrunch published 11 videos and did 114.4 million views. Brand-owned channels are the smallest group by a distance, and the two that stand out are both small: Kia America has 161,000 subscribers and its film "Deliver to Inspire" has 88.4 million views, and Snapmaker, which sells 3D printers to 65,300 subscribers, posted eight World Cup Shorts for 33.2 million views, one of them running 904 times that channel's median video. Both channels put paid media behind those numbers; a channel whose median video does twenty thousand views is not reaching 18.9 million people on its own, so read these as ad campaigns rather than audiences.

Then there is the question of who kept it. Comparing each channel's median video in the ten weeks before the tournament, during it, and in the three and a half weeks since the final separates the two outcomes cleanly. FutVibes went from 807,000 views a video to 1.0 million during the tournament and 1.2 million since, so it did not spike so much as move up a level and stay there. The Brazilian clip channel SILVADXT did the same thing, from 682,000 to 1.3 million to 1.4 million. Against that, Argentina's La Casaca Futbol went from 79,000 views a video to 301,000 during the tournament and 4,462 across the 18 videos it has posted since, and T Sports Bangladesh ran 33 times its normal during the tournament and is now back under 20,000, close to where it started. If anything the last column understates the two channels that held on, because their newest videos have had the least time to gather views.

Why it matters

The brand that won the tournament never signed a FIFA deal. Counting only the ads brands placed with creators, FotMob, a football scores app, took 21.1% of every sponsored view the World Cup generated: 45 placements, 55.4 million views, an average of 1.2 million views per placement. FIFA's entire official partner roster combined took 15.8% on the same basis, and four of the six biggest sponsoring brands had no FIFA deal at all. FotMob did it by buying the channels football fans already watch, including Futcrunch above, at prices set by their everyday numbers. The lesson for the next global event is that the winning inventory sits with creators who are priced against their ordinary reach and reachable while everyone else is negotiating with the rights holder.

Links Roundup

Quick hits from across the web

→ Twitch will let Amazon train its models on streams, VODs, clips and chat unless a streamer digs into account settings and switches it off, and within hours of Wednesday's announcement nearly 14,000 people had backed a support-forum comment demanding it be opt-in instead.

→ Alex Cooper's Unwell was valued at $500 million on Wednesday after taking its first outside investment, from Patrick Whitesell's WTSL, on the back of a podcast network whose audience is 89% female and 72% aged between 18 and 35.

→ X is retiring Creator Revenue Sharing on September 7 and replacing it the next day with Original Content Rewards, which pays on unique impressions from Premium subscribers rather than on engagement, explicitly to stop rewarding recycled posts.

→ After three and a half years without advertising, Kick has launched mid-roll ads and opened itself to brands for the first time, pitching an audience of more than 100 million users of whom 81.7% are aged 18 to 34.

→ Whalar finished moving into Accenture Song at the end of July, which puts an agency with more than $600 million of creator campaigns across 40 countries inside one of the largest consultancies in the world.

Creator on the Rise Add to your watch list

Wickens Wicked Reptiles

Exotic pets • 439K subscribers • Canada

Wickens Wicked Reptiles is Adam Wickens, a Canadian comedian who has spent seven years building YouTube's most watchable reptile channel out of long-form room tours, feeding days and rescue stories. His July video defending the hobby against a critical documentary has done 69.7 thousand views, two and a half times his typical 27.5 thousand, which is what an audience rallying behind its authority looks like. Brands noticed a while ago: 14 sponsors have booked 64 ads with him, and several, including Zbiotics and Magic Spoon, keep coming back. Exotic pets is a devoted, pet-owning audience with almost no other credible YouTube ad inventory.

Partner with the channel →

Advertiser Spotlight

Brand Strategies Worth Stealing

Skool Turned Its Customers Into Its Media Plan

Skool is a community platform: creators and coaches run paid membership groups on it, with courses, chat and a leaderboard behind one subscription, and Alex Hormozi bought into the company in 2024 and moved his own programs onto it. In the last 60 days, ThoughtLeaders' data found 12,747 videos carrying a link to a Skool community in their description, more than two hundred a day, and almost none of that visibility was bought. The links are there because the creators put them there. Every one of those videos belongs to someone who runs their own community on the platform and plugs it in every upload, from a military aviation channel with 855,000 views on one video steering viewers to the community it built with Hormozi, to a kids' drawing channel pointing its audience at its sketch club, to one business channel that posted 397 link-carrying videos in 60 days on its own.

The move worth stealing is building the advertisement into the product. A Skool customer only makes money if their own audience knows their community exists, so every customer promotes the platform every time they upload, at no cost to Skool, and the company spends its actual effort where it compounds: recruiting anchor creators like Hormozi whose communities make the platform the visible default for everyone watching them. Most brands pay for every impression they get on YouTube; Skool arranged things so its paying customers generate the impressions instead. The honest caveat is that this only works when your customer's success depends on their audience finding the product, which is true for community platforms, storefronts and link tools, and not true for most consumer goods. If your product does clear that bar, the distribution is close to free and it scales with your customers rather than your budget.

Trending on YouTube

What's going viral right now

Speed and Kai Cenat Have Been Playing Minecraft Since Friday

IShowSpeed and Kai Cenat went live on August 7 with ten days to beat all four hardcore bosses in a single run, Speed streaming it on YouTube and Cenat on Twitch. Six days in they are 26 world resets deep and still going. The ten streams on Speed's channel have done 84.0 million views between them, with day one at 12.0 million, and the marathon runs until roughly Monday. Watch day one.

MrBeast Gave Away a Mansion

"Last To Leave Mansion, Keeps It" went up on August 8 and has done 63.6 million views in five days, which makes it the second-biggest new upload on the platform this week. It is a return to the plain last-to-leave format after a run of philanthropy videos, and the reaction and breakdown channels have been on it since the weekend. Watch the video.

The "Avengers: Doomsday" Trailer Is Still Climbing

Marvel put the trailer out on July 20 and it has kept accumulating rather than fading, sitting at 62.9 million views in ThoughtLeaders' data more than three weeks later. The film is a 165-minute Russo brothers picture arriving on December 18, so film YouTube has four months of frame-by-frame reads, multiverse theories and casting arguments ahead of it. Watch the trailer.

The Biggest New Upload of the Week Came From K-Pop Again

Stray Kids released "This & That" on August 6 and it has done 65.3 million views in a week, which puts it ahead of MrBeast and everything else published on YouTube in that time. It is worth registering how routine that has become: a Korean label can name a date months out and reliably deliver a bigger audience on it than the largest creator on the platform manages with a mansion giveaway, which is not how most media plans still treat K-pop. Watch the video.

"Toxic" Released Its Trailer in Four Languages at Once

Yash's first film since KGF 2 is a Goa-set gangster drama directed by Geetu Mohandas, out on August 26. Its trailer went up on August 8 in Hindi, Tamil, Kannada and Telugu cuts on the same day, and the four have done 66.7 million views between them, with the Hindi version at 28.8 million. Releasing every language version simultaneously rather than staggering them is now standard for Indian tentpoles, and it gives four separate reaction economies the same starting gun. Watch the trailer.