YouTube finally admitted what we all know:  a view isn’t a view and a creator isn’t a creator.  When the YouTube partner program launched it was simple.  You make videos and get views.  We sell advertising and split the money with you.

But with last week’s Partner Program changes, it’s clear that YouTube values views and creators in different ways, and the simplicity is gone forever.  Why?  Because YouTube invented the creator middle class and then broke it with Shorts.  Shorts, if you remember, was built to blunt TikTok, and it did an admirable job of that.   But now YouTube has essentially split their creator economy up.  You don’t get ads unless you have sustained scale, but you still get commerce opportunities and fan monetization.  Shorts creators who don’t hit the higher 20M view threshold get “incentives” instead, including shopping bonuses and culture/trend development.  Details here are sparse, but it sounds like some of the loosey-goosey payment models the other short-form platforms have rolled out, which essentially equate to nothing.

If you missed the news, here’s a quick recap: Early next year, YouTube will raise the bar for new creators to access their Partner Program. The subscriber bar stays at 1,000 subs, but the view bar for shorts creators has doubled to 20M over a trailing 90 days, while the annual watchtime bar for long form jumped from 4,000 to 8,000 hours a year. 

So how to think about these changes?   The more I thought about it, the more I compared it to out-of-home food.  I was probably hungry at the time.  But bear with me.  Below is a standard, four-quadrant chart loved by management consultants everywhere, describing YouTube’s new world view (in an admittedly AI-generated image in a story calling out the unintended consequences of AI slop).

The biggest creators with sustainable audiences are in the upper left quadrant.  Just like an expensive chain restaurant where patrons spend hours and big bucks on a meal… and come back regularly…  their viewers spend significant time and attention with their videos, paying with their engagement.  They get the partner program, rev share, and more.  

On the lower right, Shorts creators who don’t cross the threshold of views remind me of a gas station sandwich from a Quick Stop food mart.  Instantly forgettable (unless it’s been languishing on the shelf for too long), it fills a fleeting need but probably delivers zero nutrition.  It’s an add-on to the gasoline business, not a primary business itself.

On the upper right?  The titans of fast food, with billions and billions served.  Not really any better for you, but the scale gets rewarded.

The lower left rewards lingering and attention, but the creators and content are more targeted.  Nano creators with strong communities might not make it into the partner program, nor will they not reap much in the way of rewards if they do.  But there are still ways to convert that strong engaged audience without premium and sponsorship revenue.

Sitting on top of that, YouTube now has different products for different types of creators.  Those that drive lots of attention and engagement get revshare.  If you show strong parasocial relationships and audience affinity, you can monetize with memberships and super chat/thanks/stickers. And if you can incent your fans to action, you’ll benefit from shopping and affiliate links.

There are many ramifications for creator businesses here.  First, as I’ve warned here recently, virality is dead.  Even if you get organic virality (which is less and less likely), there’s not a lot of value, except if you can snag some of those trend/culture incentives.  Whatever they are.

Second, the only creators that can really make a living wage in YouTube’s future need to live on the upper left.  YouTube is organizing and rewarding what advertisers want, which increasingly is engaged views, and trusted connections between creators and viewers.  That also happens in the lower left engaged quadrant, but not at the scale YouTube needs to support its volume business, and probably not the revenue those creators need to pay rent.

The impact of unchecked AI slop also plays a role.  Just like that gas station sandwich, or Big Mac, these forgettable, mostly short-form videos offer dopamine squirts but lack nutritional value.  Expect reduced monetization for viral shorts lacking engagement even if they are human-crafted. 

Whatever happens, there’s even mroe reason for creators to build their own viewer relationships and their own off-YouTube businesses.  Just like “Joes” has a mailing list and incentives for repeat customers, you’ll need that too.  And if you’re a gas-station sandwich, you really need to build a gas station.

Also note that this creator segmentation comes after YouTube started bucketizing viewers two years ago, based on viewing device and willingness to subscribe… with separate products and experiences for each quadrant.

I tried to map YouTube’s monetization options against this four-quadrant view.  Not sure how successful I was, but you can see the results in the chart below.  (YouTube)

Mr Beast released a video two weeks ago, which was a re-creation of an AI-generated viral video of “him” from two years ago, using the soundtrack from another viral Russian rap video.  The Beast dance turned into TikTok’s most reposted video ever, with nearly 300 million views as I write this Friday.  A few things to reflect on here.  First, the Mr Beast Rizz dance meme was likely revived in July, when the Duolingo account released its own take on the dance.  And to “prove” that his video was real, the Beast account released a second video two days later with an outtake from the video shoot.  

Even more telling, a new head of TikTok started at Mr Beast recently and has been revisiting and reviving old memes.  Needless to say, he’s had a big impact so far. 

And that’s a textbook case of how to battle the bots.  Take their inspiration and make it human.  The awareness alone was worth it, but I doubt the dollars earned paid back the time spent making the video.  It’s also an example of how quickly memes get recast, rebuilt, and recreated across the internet (see the next story).  (MrBeast)

Another creepy open-source IP is headed for the movies. Tongal is partnering with Creator Savanah Moss, a TikToker who also has a significant YouTube and Instagram presence, to make a video on Jeff the Killer.   But there’s a twist here.  Even though the IP has been created and developed by many netizens over time, Tongal says it has secured “the rights to develop a feature film based on Jeff the Killer”. 

But even if you somehow have clear ownership of a character, you still have some tough hills to climb.  I was chatting with @Diana Williams, who is building in this space, and she said copyright isn’t enough:  ”In addition to confirming who has the legal rights to a story (or character) born of or within an internet community, you have to now consider… if that community will accept your specific creative take on it.”  Reddit, as usual, has issues.  (Reddit)

TikTok A/B tested a safety feature to see how much engagement would decline if it that feature launched.  The control group included a kid who committed suicide after his feed filled up with thousands of videos of suicide, sadness, hopelessness and loneliness.  Withholding child safety features while testing business efficacy has become cornerstone evidence in lawsuits against social platforms.  It also makes me literally sick to my stomach. (No One Planned This)

Have you had a chance to watch the AI generated videos of a virtual me sharing a weekly recap of this newsletter?  They’re fun, I think, although some call them creepy.  I just love that they exist at all, and that I get to experiment with the tech.  And now I’m not the only one doing these weird AI videos.  One of my favorite analysts, @Doug Shapiro, just worked with Popcorn to turn one of his in-depth analyses into an AI-generated video featuring AI Doug.  It’s really good, clear and better than mine.  Want to see the future of AI video?  Watch his!  And mine too!  ( AI-Doug, AI-Jim) 

Watch what I do, not what I say.  That’s one of the big takeaways from new research from CreatorIQ and Influencers.Club of 5,095 creators and an unknown number of brands. (full disclosure, I am on IC’s board). 

Brands rank follower count last from a list of 8 things that drive creator selection.  But follower count has the tightest correlation with creator income across Instagram, YouTube and TikTok.

There’s a disconnect here but why is unclear.  Brands may choose creators based on fit, but then price deals based on reach.  And the study measures total creator income across all sources, not just brand payments.  Still, if you are a micro creator waiting for big brand payouts, don’t get caught up in the trust, fit and deep engagement hype.  Creator income argues against it.

Most creators aren’t waiting for revenue to catch up with rhetoric, as half have launched, or plan to launch, their own brand

Other fascinating takeaways.  72% have used AI, but mostly for brainstorming ideas and writing/editing captions.   Agentic workflows are pretty much nowhere to be found.  And Instagram edges out TikTok as creators’ best long-term bet.  

There’s also tension between what creators think their audience wants and what brands want them to make.  42% feel it overall but it jumps up to 53% for Instagrammers with over 500k followers.

One big caution with this report: it claims global insight, but 63% of respondents came from the US. And when you add in the UK, Canada and Australia you get 92% of the sample. Creators from another 90 countries participated too, averaging fewer than five per country. That’s barely more than a curling team.  I guess global = English speaking here, which is odd because Influencers.Club is based in Eastern Europe. (CreatorIQ)

That’s the conclusion of Agentio’s latest research that, unsurprisingly, shows why using Agentio is smart.  They evaluated 65,000 Meta Partnerships ads and found that brands get 19% more clicks and spend 5% less than running the same creator content on their owned channels.  Also, 45% of the winning ads looked like losers early on, so don’t cut and run until you’ve spent $1000 on testing.  And don’t boost past 36 days… after that, the cost per acquisition roughly doubles.  Agentio skips sharing how much creators earned, which is a big miss.  (Agentio)

Media Partners Asia put together an analysis of microdrama platform Crazy Maple Studio (ReelShort) that reads like sell-side equity research.  It’s super helpful if you’re evaluating or scaling micro-drama studios and platforms.  The most interesting takeaway?  ReelShort is a performance marketing company disguised as a streaming platform and spends 55% of revenue on user acquisition.  

Counter-intuitively, the report concludes that AI’s real impact on microdramas is increasing the success rate, not lowering production cost.  AI-assisted development lifts success rate from 5% to 7-8%.  That’s worth much more than the 30% production savings.  Why? Because hit shows have three monetization “windows.”  First, via paywalled access, and then by cutting viewer acquisition costs by converting the best scenes into clipped promos. Finally, franchise building: with a hit, AI enables rapid sequels that strike while the iron is hot.  The report also notes that 40 to 50% of top titles in China are already AI-generated.  The other area where AI excels?  Localization.  ReelShort is a top 10 app in Poland and Romania, for example, with almost zero paid marketing. 

The report also throws darts at Hollywood’s expensive production model. First, unlike traditional studios, microdramas resemble venture capital, where only about 1 in 20 wins… but each win pays for the entire slate and then some. Quickly churning, testing and launching beats high-quality production of a single “drama”.   Second, the top two microdrama apps (RealShort and DramaBox) sit atop most global download charts, out-downloding Netflix nearly every time.  Note, this isn’t a public download, you’ll have to “inquire” for access.  But it’s well worth reading.  (LinkedIn)

Platforms should focus on rate of growth, not overall size, to optimize revenue and viewership.  That’s the conclusion of a new study from Hong Kong University and Stanford that modelled how algorithms at advertising/subscription platforms should allocate traffic to maximize profit.  It’s not just betting on winners, as that leads to leaving up to 25% of value on the table.  The true path to profitability caps your stars and focuses on a gated curated middle-class of rising stars.  

My take?  Betting on winners is the wrong strategy, as momentum is the real path to maximizing value.  But remember, this is research borne out of a simulation, without real-world data.  It’s just a theory.  And not even a game theory. (arXiv)

🤖🤖 SIGNAL VS. SLOP

Every year I build a big presentation on the weird, creepy, and genuinely wonderful future of the creator economy. I took this year’s to SXSW – focused on how AI will change the creator economy and what human creators and the industry should do.

THE ONE PERSON 24-HOUR NEWS NETWORK: Mirage, the company powering AI-editor Captions, just released a 24-hour test of a live news channel, featuring anchors covering breaking stories, stories from wire-services and interviews with real people, including my friend @Aishwarya Srinivasan (sort of). Mirage says it’s the first 24-hour news network run entirely by AI.

It cost around $50,000 for a day of content, which is $18M a year for what feels more Max Headroom than CNN.  It’s definitely not perfect.  For example, pronunciation remains an Achilles heel, as at one point an anchor pronounced Microsoft’s Azure like it rhymed with A-Sewer.  

But forget the cost, it will continue to drop.  AI has now moved from creating clips to full-on continuous channels of content on a channel that never sleeps and needs no one…  not even a booker.  The guests were AI generated as well.  The solo creator running an entire family of channels is that much closer today. 

If one person can run a 24-hour daily news network by next year, what’s your plan to compete? (LinkedIn)

This was one of 44 different themes I presented at SXSW in March.  Look for a new theme each week as we explore the creepy, weird and ultimately hopeful future of creators in our new AI economy.  Grab the full deck free here: https://buymeacoffee.com/jlouderb/e/521396