Executive and Creator | 2x Esports and Brand Ad Agency CEO | Studying Marketing, Tech, & Content Creation Ideas in the New Media World
Joined September 2012
- Tweets 4,216
- Following 1,255
- Followers 44,881
- Likes 3,812
299 Photos and videos
Pinned Tweet
4 Aug 2025
Where to find me:
Youtube: youtube.com/devinnash
Community: discord.gg/devin
Watch: twitch.tv/devinnash
Support: patreon.com/devinnash
Consult: bit.ly/consultwithdevinnash
If my work has positively impacted you, you can find 150 more videos on my Patreon for $5.
2
121
31,674
May 8
Tapping in on the Twitch viewbotting discussion as an actual agency brand advertising executive.
This Twitch post is unlikely to convince any advertisers or streamers that the issue is improving.
Here's how to actually address this problem -
Viewbotting is an engineering and incentives problem disguised as a cultural one. There is no solution to this problem that can ever be achieved via culture change or moderation. People are debating the efficacy of the solution in Twitch’s post when the entire implementation isn’t even relevant.
Given Twitch’s track record, no one should believe Twitch will enforce this “viewer cap” change appropriately even if they did have the proper backend data (and I think they don’t.) They won’t even enforce against popular streamers who have viewbotting programs visible on their screen, streamers who roleplay bringing slaves on their broadcasts and feed them treats, or broadcasters who advocate blatent political violence, and so on.
Many streamers are on their 8th or 9th temporary bans for various TOS-breaking content. No one believes Twitch's intervention via moderation will change anything.
So again it’s a matter of engineering and incentives. I’ll break down each one. But first a productive question to ask. Why does Youtube Live - which now eclipses Twitch’s live viewership by 40% - have nearly no problems with bots and viewbotting?
Well first the engineering problem. Viewbotting gets solved by better detection, pattern recognition, IVT analysis, device fingerprinting, IP patterns, watch-time anomalies, removing off-platform embedding, and other boring backend stuff. As Dan correctly noted, these are things you can’t be public about because they’ll quickly see counterplay from the viewbot services. You just have to lock in and fix it. Google has spent billions of dollars solving this because advertisers don’t like invalid traffic. Their IVT rate now sits at a healthy 11% or so, easily beating competing live services like Tiktok (24%) and Twitch (35% .)
Twitch hasn’t taken this problem seriously. I’ve spent hundreds of thousands of dollars on Twitch, and no team member has ever communicated to me or any of the agencies I work with about their IVT rates accurately. On the contrary it often feels like they throw raw viewership numbers around in pitch decks and presentations as the REASON to go advertise there. That used to work, but money is getting smarter in digital. Brands correctly care about return on spend and engagement with real people rather than just the perception that there are viewers around.
You need both bravery and expertise in engineering to make massive changes. Twitch actually has phenomenal engineers, many who were brought on by the former CEO to develop the world’s most advanced CDN (content delivery network) from 2015-2022 before Youtube Live recently surpassed it. But I’m not sure the company has the bravery to make real, necessary changes that many commentators on this post have noted must happen. Successfully “fixing” this would represent a 30-35% global drop in viewership for a website already perceived as ailing against its competitors.
To solve the engineering side you just have to apply more of the above technical solutions and spend money and time. Communicate with advertisers, refund us in high IVT events, keep improving your CDN, etc. It’s not a moderation issue, it’s not a viewer cap issue, that’s crowd pandering and will fix nothing. Both advertisers and the community begged for years for attention on this topic, while viewbotters bled us all and collected millions of dollars. Twitch has eroded trust and should just stop making these posts to the public. Users have lost tolerance for the platform and just need to see the actual changes.
The second problem is the incentive problem. Every system to succeed on Twitch is set up based on the number of viewers you have. The entire Partnership/Affiliate pipeline is viewership based. The entire discovery funnel is viewership based. Sponsorships are calculated on CCV per $. And the entire culture of the website is “more viewers = better.” So of course everyone will do anything possible to get more viewers. This not only encourages viewerbotting but also toxic streamer culture (IRL nuisance streamers, etc) because, again, it’s all about the views.
On Youtube Live monetization rewards sustained engagement and overall channel performance. Ads, super chats, channel memberships, etc. Discovery is pushed by an interest-based recommendation algorithm. Viewers will find you because of interest alignment - not because you have the most viewers. Barely anyone viewbots there because there’s no point. Adding viewers won’t help your channel, and Google’s AI/ML will instantly and automatically suppress your channel and shut off your ads.
It is frustrating that the C-Level over at Twitch keep posting about viewbotting as if it’s going to get solved by human moderation. As if Twitch is going to magically have an amazing enforcement team with “the data” that goes after the serial offenders and returns everything to normalcy. This is all solvable, even from where Twitch is now. But you have to make the investments on the engineering/CDN side and just get to work. Everyone is waiting and hoping for these changes.
Twitch’s viewbotting epidemic is a downstream effect of what every major streamer and industry professional told them would happen for years. They didn’t implement discovery engines, they didn’t change the incentive systems, they didn’t invest aggressively against obvious botting cases, and they didn’t communicate. Now they have an existential advertising and creator crisis. You reap what you sow.
May 7
A note on our work to combat viewbotting, from CEO Dan Clancy:
There’s been a lot of discussion recently about viewbotting on Twitch, and I wanted to share an update on our enforcement efforts.
Viewbotting is bad for our business. We don't benefit from it, and we believe it harms the creator ecosystem overall. However, effectively combatting viewbotting is challenging. As we deploy updates to our real-time detection algorithms, viewbotting companies quickly respond with updates to avoid detection. Also, our detection systems must be precise to ensure that legitimate viewers are appropriately counted.
Today, we’re introducing a new enforcement type that we plan to roll out over the next few weeks. For channels identified as persistently viewbotting, we will apply a cap to the streamer’s CCV for a fixed period of time, on all of the Twitch surfaces. The cap will be based upon historical data regarding that creator’s non-viewbotted traffic. Repeated violations will result in longer penalties. Streamers will be notified when an enforcement is applied, along with the duration of the penalty, and can appeal through the appeals portal.
While streamers will be notified, we will not make a follow-on announcement when we begin issuing these enforcements, and will not publicly share details about when and where these enforcements are applied. Unfortunately, providing details simply makes it easier for companies to work around our interventions.
We believe this approach will help us make meaningful progress against viewbotting. We will continue refining our systems and expand when we apply these enforcements over time.
- Dan Clancy
50
71
535
98,662
Apr 15
If you understand this chart you understand the state of livestreaming in 2026. You can propel any influencer to high viewership regardless of content quality by tricking algorithms into featuring them. It guarantees followers and is basically pay to win.
Here's how it works:
Central Discord servers post opportunities for anyone to clip any content from a streamer in exchange for CPM ($s per 1000 views.) Typical CPMs are .50c to $2, with a minimum view limit for payout of 100,000 views.
The clippers are mostly second or third world where the USD currency exchange rate is more favorable. Platforms are usually Tiktok, Instagram, and YT Shorts, but also sometimes X. Clavicular has said payouts are sometimes as high as $30 CPM, but I don't believe that.
Because of the view limit, many of the views on these charts are free. Clavicular had 2.2 billion views on 70,000 clips, putting the average clip at 31,700 views. That would suggest most of the views are free (below the 100,000 view limit.)
I am also told a single clipper can count multiple clips towards that view limit. If that is true, it would make these campaigns much more expensive than what I state below.
Assuming 90% of the views are free, it would cost about $222,123 a month to run his campaign alone at $1 CPM. That number goes up to $666,371 if only 70% of the views are free. It's also exponentially higher if the CPMs are even remotely where Clavicular says they are (it would be in the millions per month.)
Now here's the interesting part. None of these people are running these themselves. With respect to these streamers, some of them are quite smart but none of them have the marketing prowess to run a campaign like this.
These campaigns are being run and funded by Kick themselves. To quote Clav on stream last month: "I believe it’s about a thousand clippers right now. A billion views a month. Kick has helped push me a lot with their clipping budget, over six figures a month in that campaign going towards pushing me out to new audiences."
Kick is spending MILLIONs of dollars a month promoting these streamers, and then is also paying them via their partner program as well. It seems to be totally arbitrary who gets promoted and who doesn't, a "who you know" sort of thing.
This is a different approach to Twitch's growth marketing strategy that they executed from 2012-2019ish, which was to pay streamers for exclusivity and the number of viewers they had with MGs and bonuses.
It's a much more effective strategy because it abuses algorithms. 70,000 clips in a month of Clavicular tells an algorithm "hey a ton of people are posting about this person, make sure anything he posts blows up."
You could be the most uninteresting person ever (indeed some of the people on this list are) and still get massive reach because it's not the people that are voting on your popularity with engagement. It's the algorithm pushing it no matter what because it's taught that sheer volume of posts about the same thing = that thing is important.
From one perspective this is astroturfing content creators and creating false ecosystems. From another it's a new age multi-million dollar marketing campaign from a platform to drive viewers to itself.
Wherever you stand on it, streaming is a much more "gamed" ecosystem now and will remain so. There's no real way to police this behavior since it hijacks how algorithms perceive growth and serve content.
This makes livestreaming the most "rigged game" in content creation and you will see very few broadcasters come up in the next couple years who aren't utilizing this strategy.
56
62
691
119,748
Apr 15
Apparently editing your post on this platform completely destroys its reach? Wow, great. If it doesn't pick up I'll probably delete it here, and repost it to Substack. Also will make a Youtube video with any clarifications needed.
5
1
99
8,681
Apr 10
Influencer offered $4,000 for the post ($281 CPM) and Phoebe counters $400 ($28.17 CPM) Then puts Phoebe on blast for not having enough money. I've done millions in creator advertising spends and $281 CPM is up there with the most ridiculous offers I've ever seen.
Even still Phoebe countered with a rate that was well above what the influencer's views and engagement deserve. For context a Tiktok post will usually go for $4-$10 CPM. Phoebe's counter was more than reasonable and she was very polite about it.
For a healthy ecosystem both creators AND advertisers need the deal to make sense on both ends. You'll always have a wide range of deal values and subjective considerations (engagement %, audience trust, etc) but neither side should try to extort the other. In this case the influencer was way beyond the pale.
All kinds of better things to do with $4,000. Literally impossible to get a return on advertising spend for that on 14,200 views average. I could hire ten people with signs for a week to walk around a downtown city and get a better ROAS than that.
The real mystery is why Phoebe was even doing her own influencer activations. Respect for standing on your own two feet and hustling but you get an agency to avoid exactly this. @PhoebeAdellle - I probably would open one more spot to AOR for Phia. Just saying.
25
11
656
195,739
Apr 4
Becoming a social media influencer is the most desired career in the United States and yet no one seems to want to put in the actual work to become one.
There’s this perception that it’s difficult but in reality basically no one even tries.
- Only 10% of Podcasts survive to episode 3. 90% of those quit before episode 20. You are in the top 1% of surviving podcasts if you upload just 21 episodes.
- About 85% of Youtube channels don’t upload more than 5 videos.
- About 90% of creators quit within 90 days of starting their channel, on any platform.
- In livestreaming, over 70% of streamers quit by 6 months. 90% will quit within their first year.
- 10% of X users are responsible for 92% of all tweets. (Pew Research Center)
- If you upload 20 pieces of content consistently on any platform you are past 99% of people who ever tried.
- A 26-week study (by Buffer) of 100,000 creators showed that if you post in 20 or more weeks in the 26 week window you see 450% more growth than if you post in 4 week windows or fewer, *regardless of your content quality.*
If you take 100 people interested in content creation who you think you are competing with, you are really competing with only a single person out of those 100 who might stay consistent. That person is probably not even in your niche.
Your success in content creation depends on your ability to push past fear and post consistently.
Just a reminder that you are never competing against anyone in content creation. You’re competing against yourself.
80
146
1,415
237,118
20 Oct 2025
I think most people intuitively understand extreme content is bringing down Twitch but can't place why. Twitch will NEVER be a sustainable or respected platform with the content it currently shows. the real reason why is an engineering problem. it's an issue of ad segmentation.
on Youtube/Tiktok, content classification for ad eligibility is done automatically via algorithm. it determines content can be monetized or "on" - green - "limited" - yellow - or "ineligible" - red. green is Google's entire adsense network, limited is for advertisers who opt into higher risk types of content, and red is content deemed not acceptable to any advertising audience. the guidelines are clearly outlined, just Google "Youtube Content Classification Guidelines" and click the first link.
because of proper ad segmentation (and segmented discovery queues, but that's another conversation), you can have extreme content exist in a containment field where advertisers and their users will effectively never engage with it. as an advertiser selecting for brand safe content, I can and have run hundreds of thousands of dollars on Adsense and never had my advertisements associated with brand unsafe content.
Twitch has no ad segmentation system, so if you buy inventory on Twitch it is displayed site wide. This is problematic because it means your advertisement will be shown against the lowest common denominator content. But even worse it means your CPMs will reflect that content instead of their proper category. So for example where finance CPMs should be closer to $25-35 as they are on YT, you'd get whatever the CPM of the most offensive content on Twitch is (very low) for ALL creators. Twitch artificially inflates CPMs in the past because of the "shiny new car smell" of a livestreaming platform, but competitors have moved in and normalized that market and they can't get away with that now. There's no reason to advertise on Twitch since it's so high risk and I'll get stronger ROAS on any other platform.
Twitch claims to have ad targeting, but this only means they can turn off individual streamers and categories manually. An ad bought on Twitch displays in a single ad fill pool. This is why the "Twitch Adpocalypse" I wrote about earlier this year happened - major ad buyers realized via dozens of emails informing them that their ads were being shown against extreme content on Twitch and pulled their budgets. This is why ad rates are so poor on Twitch today and haven't recovered since earlier this year.
Twitch will probably never build this classification system, for all the reasons I've written about recently. But also because it cost Youtube billions of dollars and years of drama and trial and error. Twitch doesn't have the stomach for it.
So if you understand this issue, the only solution is to blanket ban all extreme/political/controversial/low CPM content, as that will A) normalize CPMs and B) bring advertisers back.
Twitch literally can't exist as a sustainable platform without ad segmentation, ever. And as we discussed before, Twitch will never build that system, ever. Advertisers are allergic to the kind of extreme content the platform features. Even worse these people are constantly featured and front paged, since there is no true algorithmic discovery, so even if there was ad segmentation it would be difficult not to associate with it as a brand - since the platform itself is literally featuring these people.
So the only real solution is a deeply knowledgeable CEO in the advertising world who can white paper this out and show the path to profitability by banning 60-70% of the total content viewership on Twitch - with the understanding that in 2-3 years it will normalize advertising and CPM rates and create a sustainable platform. there's only like a few dozen people on Earth that understand the issue well enough to do this and they'd probably rather work on more interesting problems. that is why I said I wouldn't hold your breath for meaningful change.
20 Oct 2025
Can you explain why you feel Twitch needs to ban political content specifically to find success when it is allowed on the more successful platforms you reference here?
Is it just a feeling that Twitch can't compete pound for pound so it should lean into a singular niche more?
17
19
335
54,400
19 Oct 2025
Twitch is making the rounds again for incompetent decision making and If you've ever wondered "why isn't Twitch fixing this!?" this thread is for you. Let's clear up some misconceptions and have an honest conversation about where the company is at.
The biggest reason you see Twitch in this state is simply that the company is a ghost of its former self. Twitch is run as a demo product for IVS - the Twitch architecture they sell to companies like Kick that want to build their own streaming service.
In 2022 when most of you knew Twitch at its height, it had roughly 2500 employees. When it became apparent that the site was not sustainable, they had a 400 person layoff (March 2023) then another 400 in late 2023, then ANOTHER 500 in January 2024.
Today Twitch is a shadow of what it was and is generously valued at about $46 billion. But Amazon is a 2.27 TRILLION dollar company. That means that best case, and $46b really is best case, Twitch is about 2% of Amazon's total portfolio. This was from a Needham analyst and I think the real number is much less, but let's assume that's true.
Twitch is only mentioned a handful of times in earnings calls and financial disclosures, and never on its own. In 10 years Twitch has come up in Amazon public reports 4 times. 3 were in the Q1 earnings call in 2021, and the Q1 earnings call in 2024. They both were one sentence, and referred to Twitch as part of Amazon's advertising package. The 4th mention was a Q1 2025 Earnings Call, and was an Amazon executive mentioning Twitch as part of a "non-profitable sector." It has never had its own financial specifics listed publicly in a 10-K, meaning it's not material enough for Amazon to give it separate reporting.
In late 2024 Dan Clancy (current CEO of Twitch) said Twitch is "not profitable at this point" and that revenue was at a five-year low. This is 9 years into the companies lifecycle.
When I got into brand advertising I started with Twitch and thought I was running hot because I was doing $20,000-$50,000 influencer activations for gaming sponsorships for streamers on the platform.
Then I expanded my agencies client base, 10x'd those deal values, and realized absolutely no one cares about Twitch. It's simply too weird, too parasocial, too extreme because of dominant political streamers and drama farmers for most advertisers to look at. If you are a non-gaming brand it is a joke to advertise on Twitch, and it's because of Twitch's direct choices to platform the types of streamers it does that this is so.
And these days warring streamer communities will literally crawl email addresses of VPs of Marketing and warn them about advertising there. This happened to more than one of my clients when we focused budget there. It's unhinged and all just too much trouble to bother.
Even in the world of livestreaming Twitch has lost out to Youtube and Tiktok Live. So it's not even a primary choice for advertisers who want that inventory. Youtube Live is 50% of the market, with 13.26 billion watched hours. Tiktok Live comes next at 9.23 billion, or 14.9% of the market. Twitch is half of that at 4.35 billion, representing only 6.3% of the market. For perspective, 8 years ago Twitch was over 75% of the market. That's how far it's fallen.
So if you wonder why Twitch appears so incompetent and the laughing stock of Twitter, it's just not a relevant platform. Most of the truly great minds that worked there have left and the few that remain are marginalized and mired in corporate nonsense where any idea gets sunk into endless bureaucracy and never implemented. Most of the people left are enjoying the free meals at Twitch HQ and 6 figure salaries in San Francisco, and hoping AI doesn't clean them out of a job.
Amazon doesn't need or care to fix it. It gives them advertising exposure to gaming and 16-36 year old male demos and is a great sales pitch for IVS web services. They do not care about the content or the creators. Amazon is a consumer-goods brand, not an ad network like Google is. That is why you see Youtube as such a priority for Google - because it's ad network is integral to its success. That is also why Youtube generates tons of profit, because of all of Google's business model can feed into it. But Amazon has a much weaker ad network, and it's directed towards selling its products on Amazon. It was never, and will never be, a content brand. They just don't care about that.
If this post feels like I'm dooming on Twitch, I'm not. I actually think Twitch is pretty AI-resistant and a great platform to create on if you have a solid top-level discovery funnel that doesn't depend on it. You should never ever expect new viewers from Twitch. I also think it is not so great a loss leader for Amazon that it won't die and instead just remain a rudderless ship, with features gradually being stripped so it doesn't bleed Amazon's pockets too much. That could change in a long-term recession, but it's unlikely.
I still love the platform and watch it everyday - mostly the OG gaming creators like Lirik and CohhCarnage who I think are the lifeblood of it.
I just wish it was honest with itself and did what it does best, be a community-driven gaming platform. It's sad to see it lose its way. I don't see that improving without an extreme visionary CEO taking it on and convincing Amazon it needs serious change.
I miss what Twitch was, and I'm still adamant that banning all political content is the first step to getting it back there. I'm grateful though that there's still a lot of authentic creators and I hope they still make careers on the platform, albeit they are wise enough to diversify.
But if you've ever wondered why nothing seems to change and they make mistake after mistake, this is why. It's sometimes funny to witness how hilariously bad they mess stuff up but it's unfortunate in that it negatively impacts a lot of lives, both creators and users. I hope it changes, but I wouldn't hold your breath.
75
245
2,189
386,214
1 Oct 2025
This post has come to pass and we're quickly reaching the point where AI video is indistinguishable from human generated content.
Currently Youtube is SUPPLY constrained but it's about to become demand constrained. How will creators earn money or get discovered? Some thoughts:
This is the biggest issue we should be thinking about in AI video (besides the truth problem) and no one is talking about it.
CPM-based advertising and algorithm discovery are the two driving forces behind making content creation possible on platforms like Youtube, Tiktok, etc. They help each other by getting views on videos then monetizing those views from advertisers. And they're about to get VERY weird or break completely.
Youtube currently has about 20,000,000 videos uploaded per day. While that sounds like a lot, many are nonsense, and it is WAY less videos than the userbase has time to consume. Youtube has about 1.2 billion people who watch it daily, with over 2.5 billion monthly active users. Most people watch multiple videos for about 48 minutes a day.
On the advertising side, advertisers run ads on videos. The percentage of ads available for the number of videos available is called FILL RATE. Fill rates are hard to measure but my agency had an educated guess that they were around 60-80% for Youtube. Very healthy.
Thus there's this perception that Youtube is "saturated" but actually they don't have enough creators. Especially good or consistent ones. It is the BEST time to be a creator ever right now. There's more ads and discoverability available for new videos, to the point where Youtube will throw you a 200 view video on your home page just to try to get them discovered. Creators get paid on average about $5,000 per 1 million views, though this is pretty dependent on category, again, it's healthy.
But all this changes with AI video. The system completely breaks apart.
It's important to note we haven't truly crossed the uncanny valley yet. We're ALMOST there, and people are starting to see how realistic Sora 2 is. (imo models like Veo4 and the upcoming Chinese stuff will blow Sora 2 out of the water but that's another thread.) But videos still look a little weird. They can now fool boomers and less tech-educated people, but there's still a valley.
The valley will quickly disappear. There will be no way for a human to tell if a video is real or not. Then the barrier to long-form videos is actually compute, and that might hold off 12-15 minute videos until inference gets cheaper.
After that, it's gloves off. You will be able to upload hundreds of videos with just prompts. There will be MILLIONS more videos per day on Youtube, and the system was not built to categorize, serve, and place ads on all of those. Fill rates will drop to 5-10%, and there will essentially be no discovery because the viewer to video ratio is too skewed. Coincidentally, for other reasons, this is the problem Twitch is having right now.
As a content creator you will not be able to upload a video and rely on natural discovery, and even if your video does get views it will not be monetizable. There will be too many videos for the algorithm to serve up and too few ads to create significant density to make any sort of real money from them.
The offramp here is that platforms demonetize and deprioritize AI video in discovery. This works as long as AI videos have metadata, or are differentiable, but both of those differentiators will probably be gone in a year. And that's under the assumption the platforms even WANT to do this but we're seeing the opposite. Google is literally plugging Veo3 into its Shorts engine and OpenAI is creating a new social app for them.
These companies will lean completely into AI video for the chance at capturing the users on the prompt side and charge them with subscription models. Meanwhile creators will get spaghettified by the sheer volume of AI content competing with them.
Then it seems like everything goes back to ads. Shopify is integrating its store into GPT - OpenAI is building an ad platform - and people live and buy on these platforms. Organic discovery becomes very hard or disappears and the only way to funnel new users is traditional pay-per-click ads.
This sucks because organic content creation is this kind of accidental emergence that has produced some awesome stuff. And I'm not saying it goes away completely - because people will always create videos even if there's no money or fame in it, just like art. But in the next few years, the fundamental structure of the content creator career path will be threatened or disappear.
I've been thinking through this problem for a year since this post and I still see no offramp here. I hope someone smarter than me in a decision-making position can figure it out. Or else maybe content creation as a serious career was just for a moment in time. It will have to change or transform dramatically to survive.
1 Jan 2025
2025 will be one of the most significant years in content creation. In all my years in this business I’ve never seen a greater change agent than what’s coming. Below are some of my predictions for content creation and what creators should be thinking about and doing.
This thread will focus on video. As usual I’ll either look like a prophet or a fool but a lot of this is coming and those not ready will end up out of the game.
AI is going to change content creation in ways almost no one understands that isn’t using it daily. Many AI-generated still images are already indistinguishable from real ones, which alone has massive implications for image platforms (Instagram), photographers, and artists. You’re already seeing this debate about AI art vs. real art. Humans have an accuracy rate of about 48% when classifying faces as real or AI generated (CRJ, s41235-023-00499-6) - but this debate only exists because that gap hasn’t closed yet. In less than 2 years AI vs real art will be indistinguishable.
However this still is a tiny shadow of what’s to come.
Video and Youtube
The big leap in 2025 will be video. Most average, non-technical users (99% of viewers) will not distinguish AI channels from real ones. Youtube Shorts and Tiktok are dominated by AI voiceovers, and AI-driven channels are becoming increasingly popular (CourtroomConsequences for example with 1mil subscribers.)
To understand what will happen to Youtube, we need to talk about content creation on Youtube vs other platforms like Twitch. Right now almost anyone who uploads consistently and innovates can be successful on Youtube. But this is not so on other platforms like Twitch. Why? It comes down to the number of available viewers vs. number of videos/channels competing for that attention. Simple supply and demand. This is called a viewer ratio.
The viewer ratio on Twitch is 25.4 to 1. Meaning for every stream there are only 25 viewers available to watch it. This number is way less in practice though because most viewers prefer large broadcasts, then mid-sized broadcasts, and then finally small streams. Twitch has many problems but the largest, unsolvable one is simple supply and demand. There aren’t enough viewers available to fill up all the streams. We’ll talk about that more in a Twitch/livestreaming predictions post I’ll make later.
By contrast, Youtube has a viewer ratio of closer to 300 to 1. With 2.7 billion users, you can upload a video and if it’s not terrible, expect to get views. Youtube is top-heavy like Twitch, but because of the sheer number of users versus the relative lack of consistent content creators, if you build on Youtube in 2025, you can grow. Consistency in content creation is rare, with half of the Podcasts on Apple having 3 or less episodes, and less than 1% of Youtube channels having more than 50 videos. Most people give up. So there’s plenty of viewers to go around for those who work hard and stay with it.
All this will start changing in 2025. AI tools like Sora and Heygen will make video easy. AI influencers will start to show up everywhere. We will see an explosion of channels run by AI. In 1-3 years (and maybe even the end of 2025) you will not be able to tell if Vtuber personas are piloted by real people or an AI. A short time after that it will be the case for ALL videos. We are way closer to this than everyone thinks. Go to my Youtube channel to see videos of me speaking in perfect Chinese, AI influencers talking through products, etc. An AI influencer on Twitch currently has over 50,000 subscribers (that’s $175,000/month btw in revenue for anyone keeping score) and is in the top 100th viewership percentile at 10,867 viewers average. This AI is piloted by a skilled streamer (Vedal987), but the AI (Neuro) does just fine on her own on Youtube with music covers regularly breaking half a million views.
This is all happening right now. And today is the worst this technology will ever be. AI video and influencers will dominate in a way that will shrink the viewer ratio on all platforms. Imagine a world where millions (maybe billions) more videos are being uploaded per day because it will take no effort besides prompting to produce these videos. And this won’t be the low quality AI slop that you mostly see now. It will be LLMs trained on influencers and the content will be indistinguishable from the original creators. We are already doing this for my content on my Discord, and the advice AI gives is similar to mine, and I am an expert in my field.
Imagine a hundred, a thousand, or any number you want more trained influencers with PHD (GPT o3 current benchmarks) expertise in a field, and you will understand that most content creators on Youtube are facing an existential threat within two years. More creative entertainment, narrative, and documentary channels will come after that.
There will be pushback. People will react to seeing their favorite influencers get outcompeted in the algorithm. There will be a preference for human channels. There will be a problem verifying them. We’re entering a world where you can recreate someone like Asmongold talking about a hundred different subjects in a hundred different ways. People will not be able to tell the difference between videos made by him and those made by AI. AI influencers will masquerade as humans. Some will get outed in big controversies. It is going to get really weird.
This will be a problem for ALL information, including political leaders. Official channels notwithstanding, clips and video will be shared everywhere. Those of us in the advertising world already know this is a massive problem. People like Joe Rogan are being used to push supplements and advertise fake products. For 90% of people today, the fake Joe Rogan is indistinguishable from the real one. You cannot shut this down because for every single one you hit, ten more will spin up. Joe (and many major influencers) have addressed this on his podcast as an unsolvable problem.
As far as I can tell, Youtube (or any platform) is not ready for any of this. While official channels exist, the proliferation of clips and videos is everywhere. X is trying to solve this with verified users - the idea being that bots can’t spin up at scale to be paid users (Twitter Blue.) The bots would get shut down and it would cost the bot creators (who are usually second or third world agents) too much money to keep verifying them. This is a great idea, but it doesn’t address all problems, for example foreign agents, who have a lot of money and motivation.
Worse, there isn’t a platform incentive for Youtube to solve this. AI agents generate revenue for Youtube in the same way influencers do. They create interesting content so viewers watch ads. Viewers are already showing they don’t care if content is AI or not, they just care if it’s good. The hope is Youtube/Google chooses humans over AI for ethical reasons, and supports human content creators while suppressing AI-only channels. But this assumes Youtube has a way to differentiate the two, and as far as I can tell - aside from a manual prompt currently in the video upload section - it doesn’t. Especially as video gets indistinguishable from real video, I don’t know if Youtube even has a way to tell the two apart.
There are some early experiments to attach blockchain verification as a form of digital footprint. I believe this will work but there are way too few people thinking about this. The tsunami of fake content will hit way before these solutions are explored and adopted.
Many influencers in 2025 will start to lose viewership and not understand why. If you are a content creator, your best bet is to capture the audience now and build the largest foundation of followers/subscribers possible before the huge waves of AI competition hit. 2025 needs to be your breakout year. You need to hit 2025 harder than any other time.
Also certain types of influencers are far more protected than others. Innovators with creative sets and hard-to-duplicate storylines like Mr. Beast, Kai Cenat, and Mark Rober will be fine through the next several years. Talking heads, info content, and reacts without large existing subscriber bases are screwed, and I have no solution for you but to build now or buckle up.
2025 is going to be an absolutely crazy year. People who work with AI first will have outsized results versus people who don’t. They will be so far ahead. This is the year beyond all others to pay attention, study, and get to work.
Thanks for reading. If it makes you feel any better, I have no solution to this either and will probably be packing fries with the rest of you until the McDonalds robot steals that job too.
25
18
219
52,585
28 Sep 2025
Kai Cenat blows past the world record for most Twitch subscribers and hits 1,000,000 subs 27 days into his 30 day marathon.
The new wave of creators, Kai, Speed, Ibai, AMP and others have taken livestreaming to the next level. They create moments that truly have to be experienced while they're happening. You can't watch the VOD on Youtube and get it. You have to be part of it right then. Kai leverages the "I was here" feeling better than anyone has ever done on live.
Kai, AMP, and his team are pushing the medium of livestreaming. The last 30 days he's had marching bands, Ninja Warrior courses, Linkin Park playing in his bedroom, celebrity collaborations, helicopter rides, just insane unimaginable production quality that prepared VOD can't do much less a fully live environment.
There's a ton of people chanting for this guy's downfall, but the market always rewards quality if you push past the haters. Media history made tonight. Kai is a true pioneer of the livestreaming medium.
19
10
151
38,424
22 Sep 2025
Why the "but it isn't free to play!?" complaints? IMO f2p is a failed model for serious online-only games. Too much cheating. ARC Raiders would be a nightmare as f2p. Instead its priced competitive to Tarkov which seems reasonable for one of the most anticipated games of 2025.
22 Sep 2025
BREAKING
ARC Raiders is available for pre-order now! 🔥
Priced at $40 USD
155
40
1,211
108,123
9 Sep 2025
If you want to make it as a creative you have to embrace being cringe.
Some amount of people will always see your work as cringe. You should take all feedback and criticism seriously and in stride.
But don't ever let it stop you from putting your work out into the world.
A true loser is a person who thinks nothing, presents nothing, and stands for nothing. They've lost the courage to express their original soul and so they project that self-contempt onto you because you're trying. They'll critique your power to express yourself but never try to gain it themselves because they're terrified of the world.
Their singular goal is to shut you off, and turn you into one of them.
Don't ever let them; keep creating.
24
75
603
80,744
29 Aug 2025
Many creators don’t think properly about how people find them. Understanding funnels is the best way I know to get someone to take an action you want (view, follow, buy, etc.)
Here's some of the most valuable stuff I know about content creator discovery 🧵->
Witness this ice cream cone I drew (I’m a marketer, not an artist) with three parts. Top, middle, and bottom. Your goal is to get someone to take the action you want. So how does this happen?
Each part of the funnel operates as follows:
Top: Initial discovery. A person finds out who you are. Potential actions here are a follow, comment, subscribe, or even engagement like an RT or a like.
Middle: Interest. A person finds out more about you and your product, and makes a commitment of either their time or money. This is usually a subscription, for creators. (Twitch/Patreon).
Bottom: Buying decision. You’ve won trust or solved a problem and someone is willing to pay you for that. This is buying a product of yours, or subscribing more than one month.
I’m just going to write about top and middle funnels today, because it’s where most people go wrong.
- Best Top Platform Funnels Where People First DISCOVER You: X, Youtube, Tiktok, Instagram, Facebook. (in order of best to worst, for most creators*)
- Middle Funnels Where People Get INVESTED in You: Livestreaming Websites (Twitch etc.), Patreon, email newsletters, lead magnets (info pdfs, etc), learning resources.
The biggest mistake I see people make is trying to get discovered on a middle funnel platform, or trying to make money on a top funnel platform. You CAN do it but everything will feel difficult. This is because platforms are built to accomplish different things.
Livestreaming sites for example are NOT top funnel platforms, except for the top 10 streamers in each category. Everyone got confused because livestreaming platforms began as top funnel platforms (5 years ago) where it was easy to get discovered. But they evolved (devolved?) into mid funnel platforms where the main value is your audience getting to know you better. Yes you CAN grow on these platforms but you would be so much better off creating on a top funnel platform and then filtering viewers to your stream.
There are nameless thousands of streamers grinding hours as you read this for a few follows an hour. Don’t make me show the math on the $ per hour value here, it’s grim. Even the largest streamers actually source most of their audience from their Youtubes, not from streaming sites.
So you want a top funnel platform that’s algorithmically designed to get you discovered from the ground up. X and Youtube are top tier discovery platforms. A lot of new people will find you. They’re built with discovery in mind. However their core focus isn’t making creators money necessarily and they can be poor monetization platforms. Again you can do it, but you’d be much better off directing X and Youtube viewers further down the funnel to an actual product, or Patreon if you don’t want to build that.
Twitch/Patreon/Email are great middle funnels. People who have heard of you can learn more about you, and eventually support you with built-in subscription options there. This is one of the reasons I encourage EVERYONE to have a Patreon (or equivalent) almost no matter what you do. It’s the best mid funnel because it takes 5-12% of your money instead of 50% (Twitch) and 30-55% (Youtube) and accomplishes the same thing.
It’s always insane to me that most creators give up a substantial 40-50% of their income BEFORE TAXES to platform splits. This is a huge amount of money and you should not direct viewers to these funnels if you can avoid it.
For most creators, your funnel ends at subscribing. You’re losing a lot of potential though because you have viewers interested in you who would buy something else. So you could consider things like apparel, workshops, consulting, greater access, or white label products you like and it would 2x to 10x your income. Even for small creators this could represent a substantial increase in income over time.
I’d encourage everyone to draw out a simple MSPaint funnel for their own business. It’s a quick and easy plan that will make your entire business much more effective. Think about:
- Where are two places your potential viewer/customer spends the MOST amount of their time?
- How do customers learn and hear about you? Where are these people coming from?
- What mid-step could I create (or promote more) that would get viewers/customers more interested in what I’m doing?
Picture below here is the best funnel [currently] available for MOST creators. Lots of exceptions and it changes all the time. For example, political creators #1 funnel is Facebook, a lot of people can make it on Tiktok, etc. It’s not gospel so much as to help you build out your own thought process on this. My goal in writing this is for you to build out your unique situation in your head and make a model that’s best for you.
If this helped you and you’d like more of my writing on this, please RT this for someone else. Bookmarks also help the algo understand this is valuable. If there’s a lot of interest in writing like this I’ll keep doing them.
People told me to stop rageposting as much (I'm not gonna) but this time I wrote something useful instead. If you have something you want me to write about, tell me. There's not much of a hook or "buy this" at the end here really, but I did put a more in-depth video guide about this on Patreon. Thanks for reading.
12
36
303
20,423
23 Aug 2025
Phenomenal thread by xQc. I'm going to back it up with data and expose a bit more. This botting problem has stolen millions of dollars from legitimate creators. It's also a deliberate tactic employed by several major agencies and streamers. -> 🧵
First, aside from an agency or two, most don't intentionally view bot their talent. It's the streamers themselves. What ended up happening is the viewbotting streamers moved up in the directory. Over years of doing this, they were discovered because of Twitch's kingmaker system. So major agencies picked them up and now just look the other way. They all have the same data I have, but them ignoring the viewbotting is easy, they have plausible deniability, and it represents 30-40% of their revenue in many cases, as we'll see below.
xQc refers to 'ad packages' - these are sponsorships given to streamers. A brand agency will win a portion of yearly advertising budget from a company. Either that agency or the company will reach out to talent agencies, who will then provide their roster. The average deal will involve several talent from multiple agencies.
The deal structure is pretty simple:
- Deliverables: What the streamer needs to do for the deal. Usually: Stream for (x hours, usually 1-3), chatbot rotation timer, on screen banner rotation timer, !command, info section, sometimes CTAs and sometimes a Youtube video.
- Rate: Generally a streamer can expect to earn between $1 and $3 per CCV (concurrent viewer) per hour. Aka a 100 viewer streamer can expect a sponsorship (with these deliverables) of between $80-$120 an hour.
Rates are almost always counted with CCV as the only metric in mind. Agencies are lazy, and in all my time in the industry (since 2015 now?) I've never seen an actual formula employed for these deals outside the ones we used. The industry average rate is $1.27 (over 2000 deals we have data on executed 2010 -> 2023) but the rate trends higher for higher viewer streamers.
Top 1% streamers are rare, and agencies push up their rates. Their industry average is about $2.19 per viewer, per hour in the top 100, with many exceeding $3. So a 20,000 viewer streamer will earn about $43,800 per hour, minus 20% ($8,760) or total = $35,040.
If you read my last thread, you'll know that the problem xQc alludes to represents about 30-40% of total viewership on Twitch. Over the last 14 days, Twitch has made an 11% correction. Average viewers are down 11% (-252,770), and hours watched are down 11% (-84,930,847). Across all streams, there is -1.18 less viewers on average.
So now let's look at an example typical deal. A 10,000 viewer streamer with 30% viewbots earning $2/CCV will earn about $16,000/hour for sponsorships assuming a 20% take. Most agencies do a deal with the brand directly and then give out money to the streamers through separate contracts, and then lie and take closer to 40-50%, but that's another thread (heh.)
Anyway, the streamer with the viewbots earns the $16,000/hour. The agency earns $4,000 (20% of TDV) and everyone (except the ecosystem itself) is happy.
The same streamer without viewbots goes down to 7000 CCV. They now earn $11,200 an hour, and the agency earns $2,800. And this is assuming the rate stays consistent at $2/CCV - it often goes down.
You can see in this example, the streamer and agency lost 30% of the deal just from not viewbotting.
Twitch's policy is to NEVER ban unless they have definitive proof (bot shown on screen) of viewbotting. They state this is to prevent false positives (innocent people getting banned, or maliciously attacked.)
So now you know these numbers and that:
- This has been going on for YEARS (since at least 2017)
- It's almost impossible to get caught unless you're a complete idiot and show it on screen. (and yet some people still do!)
- The entire industry, including Twitch itself, is incentivized to let this problem walk.
- The problem is way worse than even Twitch corrected for over the last two weeks.
Now, earlier I mentioned the difference between brand agencies and talent agencies. Talent agencies have zero incentive to fix the view botting problem for the reasons above, so they just ignore that their streamers do it. But brand agencies represent brands first, and we care a LOT about how our advertising dollars are spent.
Brand agencies are incentivized to get the highest ROAS (return on advertising spend) so we are constantly looking for fraud. So when we see a 30-40% fraud rate on Twitch, that is a joke and we simply pull our budget to other sites.
For perspective, Youtube (Twitch's direct competition), has a 2025 IVT (invalid traffic rate) of 3.5%. Google Adwords is about 11% to 22% in the worst cases. However the CPMs are also a LOT better, and I pay between 40-50% less per 1,000 viewers than I do on Twitch.
So in what universe would I advertise on Twitch?
We became aware of this problem in 2021 or so and pulled our ads from Twitch to other social. In just our deal flow from 2022 -> 2025, it represents millions of dollars that would have gone to broadcasters. We put that money to great creators anyway on YT and other places. But sadly most of the money from the rest of the industry evaporated back into digital ads or traditional.
Even worse, most top brand agencies experimented with livestream ad budget over 2022 to 2025. When they all finally discovered this, they realized they got burned for hundreds of thousands of dollars with little to no ROAS, and so they left, probably never to return.
Viewbotting stole MILLIONS from legitimate creators, and pushed them down in discovery. The lack of action from Twitch, and the top streamers that do this, burnt out advertisers and quite literally held back the whole industry. A healthy advertising ecosystem with low IVT would have made Twitch look a lot more like Youtube is today, and represented tens of thousands more jobs for livestreaming creators.
So make no mistake that viewbotters are the worst sort of scum. I am grateful this topic is getting more attention, and hopefully with this context you can see why legitimate creators like xQc are so pissed off about it. It harms the entire industry including him. We are lucky to have people like him who stay honest and will talk about this when it's not popular to do so.
The reality is a few people made short term profit in exchange for the long term destruction of livestreaming as a whole.
Put simply, the viewbot problem is way worse even after this fix, people are still doing it, and even with the correction, it still won't bring advertisers back. It's absolutely a step forward and I applaud Twitch for that. But we have a long way to go to repair the damage that's been done.
Twitch has cracked down on bots in the 2-3 days and viewbotters/victims of viewbotting have been exposed. Streamers that are part of groups/orgs are seemingly being botted much more heavily. I don’t want to start witch hunts but the data is interesting. Go see for yourself
86
155
2,167
641,292
15 Aug 2025
one thing I never hear talked about with Mr. Beast is his art of asking. dude dialed 9 figure net worth people for fifteen hours straight to raise $12,000,000.
absurd levels of agency needed to do this. probably as important of a skill as his content. lesson here.
65
46
4,461
1,208,534
14 Aug 2025
lol 73,000,000 views and 781% increase in livestream viewers in 3 days. People see a catgirl aimbotting but I see a diabolical marketer.
17
9
305
33,981
14 Aug 2025
rileycs_ streamed 6 months almost every day and never broke 50 CCV. Aug. 12th peak at 1,539 viewers. millions more impressions by doubling/tripling down afterwards. prolly worth the 24 hour twitch ban.
2
1
79
9,774
14 Aug 2025
We're going to see a lot more of this. Executives will trade extra income for brand power. Also a fantastic follow up career for content creators.
12 Aug 2025
PayPal is hiring someone to build its CEO's personal brand for $236.5K/yr
interesting times we live in
16
4
140
22,752
14 Aug 2025
Notable because I don't see a lot of potential off-roading career opportunities for streamers/Youtubers. But any small to mid-level creator could apply for this and have a shot. Pretty cool to see the business world creating more jobs like this.
1
46
5,309