Faceless YouTube channels: the full research

The research was done by AI agents that open web pages. Some sites refuse them; where that happened we say so. The small numbers point to the source list at the end. Where a figure is our own sum or guess, it is marked “our estimate” and the basis is given.

The claim

These are examples of the claim, copied from our collection of 146 tweets on this scheme¹. View counts are as collected on 6 October 2026. Dates are worked out from each tweet’s ID number. The page judges the scheme, not the people who tweeted.

“If you make a faceless YouTube channel today, you could easily earn $10,000/month by February 2025.”

@AndrewBolis, 25 December 2024, 407,284 views¹

“You can make $17,292 per month with a Faceless YouTube Channel, if you have: 1. Wifi 2. Laptop 3. Two hours per day”

@AndrewBolis, 22 October 2025, 847,292 views¹

“I’m begging you: - Start a faceless YouTube channel - Don’t tell anyone you know - Upload twice a week for a year - Make $10k a month”

@Adam_DelDuca, 2 October 2026, 61,931 views¹

“how I created a $2M+ faceless youtube channel”

@grundstromleo, 7 July 2025, 276,780 views¹

One account appears twice because it wrote 17 of the 146 tweets, and those 17 hold about two thirds of all the views (5.37 million of 7.93 million)¹. Most of these tweets go on to offer a guide, a video or a training. We quote only the part that states the income.

What the scheme is

A “faceless” channel is a YouTube channel where nobody appears on camera. The videos are a voice over pictures, clips or animation. In the version sold as “YouTube automation”, software or freelancers write the script, read it aloud and assemble the video, so the owner’s job is to pick topics and upload.

The income is meant to come mainly from adverts. YouTube shows ads on videos and pays the channel a share. On ordinary (long) videos the channel gets 55% of the ad money⁵. Shorts are vertical videos of up to a few minutes, shown in a swipe feed. Ad money from the Shorts feed is pooled and split between channels by their share of views, and each channel keeps 45% of its part⁶.

So the mechanism is real. The question is how many views it takes, and how many channels get them.

The arithmetic

Step 1: the entry bar

A channel earns nothing from ads until YouTube accepts it into the YouTube Partner Program. Today that needs 1,000 subscribers, plus either 4,000 hours of watch time in the past 12 months or 10 million Shorts views in the past 90 days². Time watched in the Shorts feed does not count towards the 4,000 hours².

4,000 hours is 240,000 minutes of viewing (our sum: 4,000 x 60).

YouTube then reviews the whole channel against its rules, which it says takes about a month².

A lower tier (500 subscribers) exists, but it unlocks only fan payments and shopping features, not ad income³.

The bar is about to double. YouTube announced on 10 August 2026 that from 1 February 2027 new applicants need 8,000 watch hours or 20 million Shorts views, still with 1,000 subscribers⁴. From the same date a channel must hold 10 million Shorts views every 90 days to be paid for Shorts at all⁴. A reader starting today has under four months to qualify under the old bar. YouTube does not say whether an application still under review on 1 February is judged on the old bar, so the practical deadline may be earlier²⁴.

The same announcement says the change does not affect channels already in the programme. A channel that falls under the Shorts figure stays in and keeps its long-video ad income. YouTube also says it will offer bonuses for reaching certain milestones to channels under that figure. It has not published the amounts or the conditions, and says more details will follow⁴. We found no YouTube programme that pays a channel before it reaches the bar.

Step 2: the rate

The figure that matters is RPM: what the channel receives per 1,000 views, after YouTube’s cut. 1,000 views at an RPM of $2.30 pays $2.30. RPM counts every view, including views that showed no ad⁷. It also counts all YouTube payouts (ads, YouTube Premium, memberships, Super Chat), so ad income alone is a little lower⁷⁸. A different figure, CPM, is what advertisers pay. It is higher and it is not income⁷.

The best measurement we found comes from AIR Media-Tech, a company that provides services to creators. It measured 300 of its partner channels from their YouTube dashboards between May 2025 and May 2026⁸. “Median” means the middle channel: half receive more and half less.

	Median received per 1,000 long-video views	Middle half of channels
All 300 channels	about $2.30
Education and science (highest)	$10.22	$2.31 to $19.50
Transport (fewer than 10 channels)	$5.69	$3.17 to $9.81
Lifestyle	$2.98	$1.77 to $5.24
Entertainment	$2.43	$0.88 to $5.02
Gaming	$2.05	$0.70 to $3.62
Business and finance (fewer than 10 channels)	$2.01	$0.73 to $5.17
Kids and teens (lowest)	$0.33	$0.16 to $0.88

The median was $2.35 in November and $1.56 in January, so income moves with the season⁸. Only 53% of the median channel’s views earned anything⁸. Even in the best group, a quarter of channels received under $2.31⁸.

Limits of this data: these are established channels with 10,000 to 50 million subscribers that work with a paid services firm. They are not beginners and not specifically faceless channels. AIR says to treat the groups with fewer than 10 channels as rough⁸.

Where the viewers live changes the rate. AIR reports: “US audiences had a $14.67 CPM” against $0.74 for India⁹. That is the advertiser’s price, not income, but income follows it. AIR adds: “An English-language channel can still land near the bottom if most of its viewers watch from lower-paying countries.”⁹ We did not find where AIR’s 300 channels’ viewers live. A channel watched mostly in the United States probably receives more than $2.30, and one watched mostly in low-paying countries less. We found no measured figure for either that we could check.

Other sources give a similar range. vidIQ, which sells a YouTube tool, says most creators get $1 to $5 per 1,000 views, without giving a method¹¹. Creators interviewed by Business Insider in 2023 reported $1.61 to $29.30 for long videos, the top figure from a personal-finance creator¹⁰.

Sources disagree on finance. Blogs often repeat “$12 to $25”, but no page we opened had data behind it, and AIR’s measured median is $2.01 from a small sample⁸. The true rate for finance channels is unsettled.

Shorts pay far less. In AIR’s data, “Shorts RPM ranged from $0.02 to $1.48 per 1,000 views, depending on the niche”, which was 3% to 14% of the long-video rate in most groups⁹. Creators interviewed by Business Insider reported $0.04 to $0.06 per 1,000, and one received $383.13 for 7 million Shorts views¹⁰.

Step 3: the views needed

Views needed each month = income wanted ÷ rate x 1,000. These are our sums from the rates above.

Income wanted	Rate per 1,000 views	Views needed every month
$10,000	$2.30 (median)⁸	about 4.3 million
$10,000	$3.00 (the rate one tweet uses)¹	about 3.3 million
$10,000	$5.00 (the rate one of the tweet authors uses, see below)⁴²	2 million
$10,000	$10.22 (best group’s median)⁸	about 980,000
$10,000	$1.56 (January median)⁸	about 6.4 million
$10,000	$1.48 (Shorts, best group)⁹	about 6.8 million
$10,000	$0.05 (Shorts, creators’ reports)¹⁰	200 million
$17,292	$2.30	about 7.5 million
$1,000	$2.30	about 435,000
$1,000	$10.22	about 98,000

The “1 to 4 million” we use as the short form is the span from the best group’s median to the overall median.

Another tweet says a channel should earn “$350/day” (link)¹. That is about $10,650 a month (350 x 365 ÷ 12), so it needs about the same as the $10,000 rows.

The text of the “$2M+” tweet does not say what the $2M counts: ad income, all income, or the channel’s value. We did not watch the video attached to it, which may say. If it were ad income alone at the median rate, it would take about 870 million views in total (2,000,000 ÷ 2.30 x 1,000).

These views are needed every month, not once. The sums are before costs and tax.

People who promote the method give similar figures when they do the sum. The author of one of the example tweets wrote in a newsletter on 10 June 2026: “Let’s say your channel has a $5 RPM” and “To make $10,000 per month, you’d need around 2 million views per month”⁴². Another creator’s video gives 2 to 3 million views a month; we saw that only in a third-party summary²⁶.

Step 4: how many channels get there

No source we found says what share of channels reach 1 million or 4 million views a month. The nearest evidence:

What regular uploading does. This is the best figure we have that takes effort into account. vidIQ, which sells a YouTube tool, followed 32,926,088 channels in its index that had under 1,000 subscribers for 360 days²⁹. Overall, 500,409 of them reached 1,000 subscribers, which is 1.52%²⁹. The share was 0.20% for channels that added no videos, 1.86% for 1 to 11 videos, 8.65% for 12 to 51, and 30.77% for 52 or more (271,792 of 883,403 channels)²⁹. So adding many videos raises the odds many times over, and about seven in ten of those channels still fall short. Only 2.7% of the channels added 52 or more videos (883,403 of 32.9 million, our division), so keeping up that output is itself rare²⁹. Limits: this counts subscribers only, not watch hours or income. vidIQ says it shows a link, not a cause, and that its index is not a random sample of YouTube²⁹. It does not separate Shorts from long videos, and it is not about faceless channels in particular²⁹. Correction, 7 October 2026: an earlier version of this page had these figures second-hand and described 52 or more videos as “weekly or more often”. We have now read the study. All four figures are confirmed. vidIQ says 52 videos in 360 days does not show a weekly schedule, so we removed that wording²⁹. The overall 1.52%, which we had removed as unconfirmed, is confirmed and restored²⁹.
What a committed new channel’s first videos get. vidIQ tracked 137,355 channels created from January to June 2025 that posted at least ten long videos in their first 90 days, and counted views 30 days after each upload⁵⁴. The median first video got 180 views, the fifth 211 and the tenth 272⁵⁴. The median channel’s best video of the ten got 1,948 views⁵⁴. 57.8% of these channels had at least one video reach 1,000 views, and 30.8% had one reach 10,000⁵⁴. For channels posting Shorts (367,243 of them), the median first Short got 1,362 views and 48.0% had one reach 10,000⁵⁴. Limits: channels that stopped before ten videos are left out, so this is the committed group and it flatters the picture. It is not about faceless channels in particular⁵⁴.
New, active channels in one subject. vidIQ followed 36,163 gaming channels that had under 1,000 subscribers on 20 September 2025 for 360 days⁵³. 16.16% of ordinary gaming channels, 26.09% of Minecraft channels and 42.32% of Roblox channels ended with 1,000 or more subscribers⁵³. 1.30%, 2.73% and 5.22% ended with 10,000 or more⁵³. The median gain was still small: 23, 78 and 268 subscribers⁵³. So a crowded subject does not by itself shut newcomers out. Limits: gaming only, which is one of the lower-paying groups in AIR’s data⁸, and subscribers are not income.
Most videos get almost no views. Researchers took a random sample of 10,016 YouTube videos¹². The paper reports a median of 35 views, 86.93% of videos under 1,000 views, and says of videos with 10,000 or more views: “only 3.67% of our sample had that many views or more, but they account for 93.61% of total views across the sample”¹². One of the authors gives rounder figures on his blog: “the median YouTube video has 39 views”, and videos with 10,000 or more views “are roughly 4% of our data set, though they represent the lion’s share of views”¹³. This sample covers all of YouTube, including home videos, so it says little about someone working to build a channel. (An earlier version of this page could not open the paper and used only the blog. The paper opened on 7 October 2026 and its figures are confirmed.)
A newer, larger count says the same. vidIQ counted 38,229,639 long videos published in its index from 15 August to 5 September 2026⁵⁵. The median video got 234 views in its first week⁵⁵. 70.1% got under 1,000 and 8.5% got 10,000 or more⁵⁵. The top 1% of videos took 54.1% of all first-week views, the next 9% took 34.6%, and the bottom half took 0.5%⁵⁵. This is not adjusted for channel size and does not identify faceless or AI videos⁵⁵.
Few channels grow large. In vidIQ’s index of 61.2 million channels, 7.9% had 10,000 subscribers, 1.3% had 100,000 and 0.13% had 1 million¹⁴. The index covers only channels with at least one subscriber, and about 40% of them already have 1,000, so the real shares across all channels are much lower¹⁴. Growth after the first milestone is slow: among channels that uploaded in the last 90 days, the median channel with 1,000 to 9,999 subscribers gains 13 subscribers a month¹⁴. (Read at source on 7 October 2026; the figures we had second-hand are confirmed.)
How crowded it is. YouTube’s press page says: “On average, there are over 20 million videos uploaded daily to YouTube.”⁵⁸ The page is undated and gives no trend.
About 3 million channels are paid at all. YouTube said in August 2026 that over 3 million creators are in the Partner Program⁴.
The average payout is modest. YouTube says it paid over $100 billion in four years to creators, artists and media companies¹⁵. Our estimate: $25 billion a year ÷ 3 million = about $8,300 a year, or about $700 a month. So the average across everyone paid cannot be more than about $700 a month. The true average for creators is lower, because the total includes music labels and media companies¹⁵¹⁶. The sum is rough: it sets four years of payments against the 2026 number of members. An average also does not limit what one channel can earn.
Earnings are concentrated. Among AIR’s established channels, “the top 10% of channels earned 58% of all ad revenue” and the bottom half shared 4%⁹. The top quarter took 82%⁹. So the typical channel gets far less than the average.

Step 5: the costs

Costs start in the first month. Ad income is zero until the channel is accepted.

AI tools. On the sellers’ own price pages: ElevenLabs (voice) $6, $22, $99 or $299 a month (the $22 plan is $11 for the first month); invideo (video) EUR 20, 36 or 75 a month when billed yearly; HeyGen (video) $29, $49 or $149 a month²⁰. Our estimate, adding one plan of each kind with the euro at roughly $1.15: about $75 a month on the entry plans ($22 + $23 + $29) and about $190 on the middle plans ($99 + $41 + $49). At the $2.30 median, a $75 bill needs about 33,000 views a month to break even, and a $190 bill about 83,000 (bill ÷ 2.30 x 1,000).
Freelancers. Listings on Fiverr for a complete faceless video (script, voice, editing) run from about $15 for 1 to 3 minutes to $100 to $200 per video²¹. Two videos a week, which one example tweet prescribes¹, is about 8.7 videos a month. That is about $130 a month at $15 a video, and about $870 to $1,700 at $100 to $200 (our multiplication). Fiverr refused us, so these prices come from a search summary and cover short videos only.
Research tools and stock footage. These are optional; YouTube requires none of them. On the sellers’ own pages: vidIQ (channel research) is free, $16.58 or $39 a month when billed yearly⁶⁴. TubeLab (subject research) is $29, $39 or $100 a month⁶³. Storyblocks (stock footage) is EUR 18, 25 or 35 a month billed yearly; the page showed euros and prices may differ by country⁶⁴. Our estimate: adding one research tool ($17 to $39), TubeLab ($29 to $39) and stock footage (about EUR 18 to 25) to the AI tools above raises the do-it-yourself bill from about $75 to $190 a month to roughly $140 to $300 a month.

What the tweet text does not say

Of the 30 most-viewed tweets, 3 give a rate per 1,000 views: $3, “$4-$6” and “$10-$15”¹. In the tweet text we collected, none gives the subscriber and watch-hour bar in numbers, any cost, or the number of views the income needs¹. We did not read the images, videos, threads or guides attached to them. The rates are within the measured range. The piece we did not see in the tweet text is the 1 to 4 million views a month.

What people who tried it report

We found no study of what faceless channels earn as a group. What exists is a handful of individual accounts, most of them self-reported, and a set of channels listed for sale, which are under “The upside” below.

One success was checked by a journalist. Fortune reported in December 2025 that Adavia Davis told it his network of channels “generates roughly $40,000 to $60,000 a month in revenue”²³. Fortune says it “reviewed screenshots” of his dashboards and recent payout records²³. His videos are almost entirely AI-made, with software built by his business partner. Fortune reports videos up to six hours long “costing as little as $60 to produce”, operating costs of “about $6,500 per month”, mostly staff, and margins of “85% to 89%”²³. Fortune puts the total at about $700,000 a year before costs, from about 2 million views a day across five active channels, with about two hours of his own oversight a day²³. He is not a beginner: Fortune says he had been building the business for about five years before the article²³. He had run shows on TikTok, Instagram and Snapchat, employs small salaried teams, and has built and sold channels of roughly 400,000 to just over 1 million subscribers²³. Fortune puts his position down partly to timing: “Davis has had something of a first-mover advantage, given how early he was to spot the arbitrage opportunity and also his long-developed intuition for the sort of video that performs well.”²³ The income was checked by the reporter; the costs and margins are his own statement²³. He told Fortune he offers an online course; we did not find its price²³. Fortune also reports his account of what happened after he posted a promotional video for the software showing his method: “Within days, Davis said that he saw scores of copycats posting similar videos, crowding out the niche that he had built and monopolized until then.”²³ Fortune reports that he estimates individual creators “have until around 2027 to meaningfully profit from AI-generated long-form YouTube content”²³.

One new channel is listed for sale with five-figure months. A faceless history documentary channel started in August 2025 is listed on the marketplace Flippa, which marks it as vetted. The listing shows $180,049.90 in total revenue on about 9 million views, monthly revenue between $15,050.81 and $34,307.70 in recent months, and a rate that “exceeds $20” per 1,000 views³⁸. That is the strongest single case against using $2.30 for every channel. It is unsold, the asking price was cut from $490,000 to $395,000, and the title says $235,000 and 11 million views while the body says about $180,000 and 9 million³⁸. When we re-read the listing on 7 October 2026 it was still open and unsold at $395,000. It then showed $19,752 a month at a 97% margin, with revenue, main expenses and traffic marked as verified by Flippa³⁸. We did not identify the channel. More channels listed for sale, including ones marked sold, are under “The upside” below.

The detailed accounts look different from the tweets.

A New York Times report from August 2022 followed a man who spent about $15,000 on courses, a mentor and freelancers. He is quoted: “I’ve tried three courses and one expert on the side, and the only thing I got out of it was an empty wallet.”²² We read this in a republished copy, not on nytimes.com.
A creator who ran faceless channels for two years reports spending $36,000 to earn $45,000 from ads on a channel with 10 million views, about $9,000 net for over 2,000 hours. He says another channel with 64 million views earned $6,000²⁴. The same summary says he earns far more, $50,000 to $125,000 a month, from sponsors, affiliate links, products and coaching than from ads²⁴. An earlier video of his reports a channel with 12 million views that was never accepted for ad income and was sold for about $3,000²⁵. We read both through third-party summaries. The summaries disagree on one channel’s earnings ($45,000 against over $110,000), and we could not settle it²⁵.
A 2022 personal write-up reports $23,000 earned in four months from outsourced videos costing $50 to $70 each. The channel was then taken down over copyright²⁷.
A video by a creator who, according to the summary we read, sells a course says outsourcing costs “$10 to $20,000 before your channel becomes profitable” and that most people who start that way earn under $500 a month. He cites his own community of over 11,000 creators and a 5% to 10% success rate, which we cannot check²⁶. This too is from a third-party summary.
A writer reports that his faceless channel has earned “$10,000 and counting” from over 2.3 million views and 6,600 subscribers⁶⁵. He says he started in November 2024, spends 4 to 6 hours on each video himself with software he already owned, and that growth came from one video, around his fifteenth: “One video in particular got well over 500,000 views and to date has made me almost $2000”⁶⁵. This is his own statement and is not checked. He sells his own course and uses affiliate links⁶⁵. The rate it implies, about $4.30 per 1,000 views (our division), is inside the measured range⁸.
The Hollywood Reporter reported in June 2026 on an operator of faceless channels who, it says, lost $250,000 a month in revenue in early 2025 when YouTube shut down his channels in a copyright dispute⁵². The article says he now runs six faceless channels⁵². We read the article again on 7 October 2026 and the wording is confirmed. It does not say how the dispute ended or that the reporter checked payment records, so we treat the figure as his own statement. It is the largest income figure we found.

Views at the first milestone. vidIQ looked at 1,027,773 channels with 1,000 to 1,100 subscribers, in an analysis it labels May 2026²⁸. 911,166 of them (89%) got under 5,000 views a month, 73,492 (7.2%) got 5,000 to 20,000, and 43,115 (4.2%) got more than 20,000²⁸. An earlier version of this page had these shares second-hand. We read the page on 7 October 2026 and they are confirmed.

What the platform allows now

Quotes in this section are from YouTube’s own pages, read on 7 October 2026, unless marked.

Faceless is allowed. YouTube’s rules do not require a presenter. They also allow AI as a tool: “If you use automated tools or templates to help create your content, the final product must still demonstrate your creative vision and provide educational or entertainment value.”¹⁷

The mass-produced version is not paid. Content that earns ad money must “Not be mass-produced, generic, repetitive, or manipulative.”¹⁷ The page lists as not eligible:

“AI-generated content made with generic or unoriginal templates giving the impression of mass production”¹⁷
“Image slideshows, templated storylines, or scrolling text”¹⁷
“Content that exclusively features readings of other materials you did not originally create, like text from websites or news feeds”¹⁷
“AI-generated podcast hosts offering financial guidance, investment tips, or wealth management advice”¹⁷

The rule applies to the whole channel¹⁷. YouTube renamed it from “repetitious content” to “inauthentic content” on 15 July 2025 and said such content “has always been ineligible for monetization”¹⁷. At the time, a YouTube reply quoted in the trade press said “channels that use AI in their content remain eligible to monetize”³⁵. In July 2026 YouTube’s head of trust and safety explained the rules again and, as reported, said he was not announcing a new policy³⁴. Blogs that describe a new ban in July 2026 do not match what we read on YouTube’s pages.

At volume, it can end the channel. YouTube’s spam policy gives this example of what is not allowed: “Channels that use the exact same background music and repetitive AI generated imagery across many videos, with each video reading out an AI-generated script.” The stated penalty: “we may suspend your monetization or terminate your channel or account”¹⁸.

AI content must be labelled. YouTube says labelling “won’t limit a video’s audience or impact its eligibility to earn money”, and that creators who repeatedly do not label may be suspended from the Partner Program¹⁹.

Enforcement on record.

In December 2025 YouTube terminated two channels of AI-made film trailers with over a billion views between them. Its statement said they had been readmitted to the Partner Program and then “reverted to clear violations of our spam and misleading metadata policies” (metadata means titles, descriptions and tags)³¹.
Kapwing, a video-tool company, found that 16 of the 100 most-subscribed AI channels it tracked had been deleted (11) or emptied (5), as reported by Net Influencer on 29 January 2026³². Tubefilter the same day cited Kapwing for “at least 18”³³. Both figures are Kapwing’s, not YouTube’s.
YouTube’s chief executive wrote in January 2026: “To reduce the spread of low quality AI content, we’re actively building on our established systems”¹⁵.
A trade report says YouTube reviewed hundreds of social-media complaints about terminations in November 2025 and upheld the vast majority³⁶.
The Hollywood Reporter’s June 2026 article quotes a creator with 1.7 million subscribers on channels in his own subject: “The people who do the same content as me without their face in it, most of them are getting demonetized” (losing their ad income)⁵². That is one person’s observation, not a count. The article also reports that some operators are hiring on-camera hosts for this reason, and that the most successful faceless channels are narrow educational ones⁵².

Enforcement is partial. Tubefilter noted that AI channels still appeared in its top viewing charts³³.

AI-made video is also the competition. Kapwing opened a new YouTube account and classified the first 500 Shorts it was shown: 21% were AI-made low-quality videos, and 165 (33%) were what it calls “brainrot”³⁰. That is Shorts only, one account, with data as of October 2025³⁰.

Several tweets in our collection describe copying an existing channel’s format with AI; some add that a full video then takes under five minutes to make¹. The tweet texts are cut off in our copy, so we do not know the rest of the method and cannot say whether it would break these rules.

Who makes money from it

In our 146 tweets: 95 lead the reader to something: a guide, course, tool, service or community. The other 51 are labelled as showing nothing on offer. The 95 hold about 90% of the views (7.12 million of 7.93 million)¹. This is our count from labels that are roughly 88% to 90% accurate, so treat it as approximate.

The biggest route is “like, comment and follow, and I’ll send you a free guide”: 40 tweets with 6.19 million views, 78% of the total¹. One account, @AndrewBolis, wrote 17 of the 146 tweets in our collection and has 5.37 million of their views¹. Nine of those tweets say the guide usually costs $73, $83, $87 or $91¹. His website lists a “Faceless YouTube Money Guide” among its free guides, linking to a sign-up page we could not open; the website itself shows no price³⁹. We did not find a page that sells the guide at those prices. His sign-up page refused our tools and we did not see the guide or what is sent to people who sign up, so this may be a gap in our search and not a fact about the guide.

Courses and coaching. Prices found run from about $500 to $7,800; the top figure is from one unverified customer review:

The 2022 New York Times report cites courses at $500, $995 and $1,500. Its highest earner said he made $1.4 million in 2021, including from his own courses and services. The article says: “The key was his $995 course, responsible for 70 percent of his income.”²²
One account in our tweets, @wannercashcow (example), runs a free community of about 4,500 members that leads to a paid programme, the WannerCashCow Accelerator, with no price shown⁴⁰. Its sales page says: “Disclaimer: this is not easy. It takes time, dedication, and especially consistency.” The page shows individual students’ results and says it has helped more than 2,000 students, but gives no share of students who reach those results. Its refund policy says: “All purchases from CashCowConsulting B.V. are final and non-refundable.” It says buyers agree to immediate digital access and that the 14-day right of withdrawal under Dutch law ends once access is given⁴⁰.
On Trustpilot, the programme Automate Channels is rated 4.6 of 5 from 42 reviews, 95% of them five-star⁴¹. One reviewer, writing on 21 March 2024, says they paid $7,800 for the course and were refused a refund; we could not check either statement, and no reply from the company was shown⁴¹. Trustpilot notes that the company “hasn’t invited customers recently, so reviews may not be representative”⁴¹. Reviews are unverified in both directions.

What one tweet author says about ad income. The author of one example tweet (“Upload twice a week for a year - Make $10k a month”) gives more detail in his newsletter. We collected only the first tweet of his thread and do not know what the rest says. He writes that “ad revenue alone is usually the slowest way to get there”, “I’d build a channel around an offer” (a product of your own to sell), and “I definitely would not start by relying only on ad revenue”⁴². The post recommends several paid tools (TubeLab, vidIQ, TubeGen, VidRush). We found no paid course of his and could not read his shop page, so he is quoted here for what he says about ad income, not as a course seller.

Tools. TubeLab, whose accounts posted three of the tweets, sells subscriptions in three plans¹. Its pricing page shows $29, $39 (shown as reduced from $49) and $100 a month⁶³. An earlier version of this page could not read the prices. A product called TubeGen AI is sold at tubegen.ai for $149, $297 or $849 a month⁶². Fortune names TubeGen as the software built by the business partner of the operator in its article, who posted a promotional video for it²³. The website names only a first name and does not mention either man, so we have not confirmed that it is the same product⁶². If it is the same product, the partner who built the software also sells it to the public; we do not know whether Mr Davis earns from those sales. ElevenLabs, named in several tweets, pays people who refer subscribers “up to 22% in commissions over 12 months”⁴³. We did not open the links in the tweets, so we do not know whether any is a referral link.

Regulators. We found no court or regulator action against any seller named on this page, or against any other seller of faceless-YouTube courses or tools; we searched US federal sources only. The nearest case, FTC v. Growth Cave, concerned a different product (a “business opportunity” package that the FTC’s complaint says was sold through YouTube ads) and ended in settlement orders entered by the court in January 2026, without a ruling on the allegations⁴⁴⁴⁵. It is not evidence about this method. These cases concern the companies named in them only. We found no regulator action against any account quoted or named on this page, and we do not suggest any connection.

The upside

Some faceless channels do pay well, and a few pay very well. This section sets out the best documented cases. Each is marked by who checked it and by whether the person sells something on the subject. None of it changes the base rates in Step 4: these are the channels that worked.

Channels checked by a broker and put up for sale

This is the strongest evidence of success we found. Empire Flippers is a broker of online businesses. It says it checks a seller’s revenue, traffic and expenses before it lists a business⁴⁸. Its public listings data, read on 7 October 2026, showed 50 businesses with YouTube income, first listed between June 2024 and September 2026⁴⁶. 38 of them earn from YouTube only⁴⁶. 20 of those 38 describe themselves as faceless, on our keyword count of the listing summaries, and 10 of the 20 are marked sold⁴⁶.

For those 20 faceless channels:

The median net profit is $4,423 a month. “Net” means after the costs the seller reports. The range is $1,184 to $22,249 a month⁴⁶.
4 of the 20 net $10,000 a month or more⁴⁶.
The median owner reports about 7.5 hours a week in the broker’s data field. Listing texts sometimes give more: the food channel’s says 10 to 20⁴⁶⁴⁸.
The costs the sellers report run from $17 to $3,144 a month, with a median of $731⁴⁶.
The median gap between the first YouTube payment and the listing is about 13 months⁴⁶.

Three of them in detail:

Channel	Timeline	Net profit a month	Who makes the videos	Listed at	Status
Niche news⁴⁷	Started early 2024, first paid June 2024	$21,172, averaged over 10 months	Six freelancers; owner about 10 hours a week; 1 to 2 videos a week	$635,149	Sold; final price not shown
Cars, information videos⁴⁹	Launched July 2024, first paid October 2024; 340,000 subscribers, over 280 videos	$22,249, averaged over 12 months	Seven outsourced workers; owner 5 to 10 hours a week	$467,235	Unsold since September 2025
Food⁴⁸	First paid July 2025; 87,000 subscribers and 15.4 million views in 13 months	$13,060, averaged over 12 months	A remote team; owner writes scripts with AI help, 10 to 20 hours a week	$91,423	Sold; final price not shown

How to read this:

These are winners, not a sample. About 50 YouTube businesses reached one broker in 27 months. They show what success looks like. They say nothing about how likely it is.
The check is the broker’s. We did not see the records and did not identify any channel. “Sold” shows a sale took place, not the final price. All prices here are asking prices.
The sellers are anonymous. We saw no course on offer in the listings. One seller says he plans to consult for other channels using his process documents⁵⁰.
The income is priced as fragile. Asking prices run from 7 to 39 months of profit⁴⁶. The food channel was listed at about 7 months of its profit (our division: $91,423 ÷ $13,060)⁴⁸. Our reading is that sellers and buyers do not count on this income lasting. The broker itself says YouTube’s rule against reused content “is the most common reason faceless channels lose monetization”³⁷. Several of the 20 listings say they rely on “fair use” of other people’s footage, and one seller had a channel removed and restored on appeal⁴⁶.
A general caution. This is from the broker’s buyer guide and is not about any listing here: “Revenue screenshots can be fabricated.”³⁷

Channels on another marketplace

Flippa is a marketplace for online businesses. Both listings below are shown as managed by a Flippa broker. It marks revenue, main expenses and traffic as verified on both listings below; we did not check that label.

The history documentary channel described earlier, started in August 2025, shows $19,752 a month and is unsold at $395,000³⁸.
A celebrity and mystery channel started in February 2025 shows about $913,731 in revenue since it began and about $895,000 in the past year, about $52,826 profit a month, 115.4 million views and a rate above $10 per 1,000 views, with a production team included in the sale⁵¹. It is unsold at $550,000, which is under a year of the stated profit⁵¹. The listing gives monthly revenue of $36,853 to $70,267⁵¹. $895,000 over twelve months averages about $74,600 a month (our division), which is above that range, so treat the figures with care⁵¹. The seller is anonymous. Correction, 7 October 2026: an earlier version of this page gave $913,731 as the past year’s revenue. The listing gives it as revenue since the channel began⁵¹.

One operator whose records a journalist reviewed

Fortune reviewed screenshots of Adavia Davis’s dashboards and recent payout records, and reports about $700,000 a year before costs across five active channels²³. This is a reporter’s review, not an audit. He says his videos get about 2 million views a day²³. The videos are almost entirely AI-made²³. It took about five years of building and selling channels to get there²³. He told Fortune he offers an online course; we did not find its price²³. The details are under “What people who tried it report”.

People’s own statements

None of these was checked by a third party.

An operator who told the Hollywood Reporter he was making $250,000 a month from faceless channels until YouTube shut them in early 2025⁵². We did not confirm whether he sells anything on the subject.
A course seller who said he made $1.4 million in 2021, with 70 percent of his income from his $995 course²².
A creator who reports $45,000 in ad income on $36,000 of spending over two years, and $50,000 to $125,000 a month from sponsors, affiliate links, products and coaching²⁴. According to the summary we read, he sells coaching²⁴.
A writer who reports $10,000 in total from 2.3 million views, doing the work himself⁶⁵. He sells a course.
A 2022 write-up reporting $23,000 in four months before the channel was taken down over copyright²⁷.

Kapwing also estimates large yearly ad income for the biggest AI channels. Those are estimates from view counts, not payouts³⁰.

What the top tenth and the top hundredth get

No source gives dollar earnings for the top tenth of faceless channels. These are the nearest measures.

Views per video. Across 38.2 million long videos, the median got 234 views in its first week. A video in the top 10% got 8,154 or more, in the top 1% 109,778 or more, and in the top 0.1% 663,099 or more⁵⁵. At the $2.30 median rate, 8,154 views pay about $19 and 109,778 views pay about $250 (our estimate)⁵⁵⁸.
Subscribers. 7.9% of indexed channels have 10,000 subscribers, 1.3% have 100,000 and 0.13% have 1 million¹⁴.
Views at the first milestone. The top 4.2% of channels at 1,000 subscribers get 20,000 or more views a month²⁸. vidIQ estimates their ad income at $27 to $503 a month, depending on subject²⁸. Finance and business channels in that group have a median near 61,000 views a month and an estimated $251 to $503²⁸. These are vidIQ’s estimates, not payouts.
Share of the money. Among AIR’s 300 established channels, the top 10% took 58% of ad revenue, the top quarter 82% and the top half 96%⁹. AIR does not publish the dollar amounts.
Six-figure channels. A trade report says YouTube’s 2025 US impact report found the number of US channels earning six figures or more rose more than 7% in a year⁵⁷. No count is given, and we did not find the sentence in YouTube’s own post⁵⁷.

A realistic good result in the first year or two (our estimate)

The common outcome. No ad income in the first year: 1.52% of small channels reach even 1,000 subscribers in 360 days²⁹. At 1,000 subscribers, the common case is under 5,000 views a month²⁸. At the median rate that is under about $12 a month (5,000 ÷ 1,000 x $2.30), and usually not yet enough watch hours to be paid at all. The top 4.2% at that stage, with 20,000 views or more, would get about $46 or more a month at the median rate²⁸⁸.

A good first year. For a competent person, a good result is being accepted into the Partner Program and earning tens of dollars to a few hundred dollars a month from ads. That is roughly the top 4% of the channels that reach 1,000 subscribers, and most channels do not reach 1,000 within a year²⁸²⁹.

A good second year. A good outcome would be 100,000 long-video views a month: about $230 at the median rate and about $1,000 at the best group’s rate⁸. What share of channels reach that is unknown.

When it really works. The broker-checked channels suggest a pattern. They were listed for sale about a year after their first payment. They net between about $1,200 and $22,000 a month, with a median of $4,423⁴⁶. No source gives the share of starters who reach that.

What it takes to compete

Two routes, with different risks

Almost entirely AI-made channels have earned large sums, and some still do. The Fortune case is one; a reporter reviewed screenshots of its dashboards and payout records²³. YouTube’s rules do not pay AI content that is mass-produced or templated, and channels of that kind have been terminated¹⁷¹⁸³¹³². The Fortune article, which we read again on 7 October 2026, reports no action by YouTube against the channels in its case, and we do not know how YouTube judges them. The general point is that on this route income can stop without notice.

Original work is what the rules protect. That is not the same as saying it pays. It means that if such a channel does find viewers, the rules allow the income to continue¹⁷.

What the channels that succeed have

Videos that break out, not just many videos. $1,000 a month at the median rate needs about 435,000 long-video views a month (Step 3). A committed new channel’s typical long video gets 180 to 272 views in its first 30 days, and its best of ten gets 1,948⁵⁴. So typical videos cannot reach the target. It takes repeated videos in roughly the top 8.5%, those with 10,000 or more views in their first week⁵⁵. About three in ten committed new channels get one video past 10,000 views within 30 days in their first ten⁵⁴. “Committed” here means channels that posted at least ten long videos in their first 90 days⁵⁴. This is our estimate from those figures. The line between the median and the paid group is drawn by breakout videos, not by upload count, tools or subscribers.
A subject and format that pay well per view. The high-rate cases we found are long history, education or science videos. Fortune’s top earner makes history videos up to six hours long²³. The Flippa channel makes history documentaries at over $20 per 1,000 views³⁸. AIR’s best group is education and science, at a median of $10.22⁸. Among the 20 broker-checked channels, the subjects are information videos: vehicles appear seven times (five on cars, one on trucks, one on motorbikes), then history, documentary, business and niche news⁴⁶. Ten of the 20 listings describe their videos as long-form⁴⁶. Not all are long: the trucking channel gives an average video length of 6 minutes 25 seconds, and one channel posts about seven Shorts and one long video a week⁴⁶⁵⁰. One of them, a trucking channel with US viewers over 55, states a measured $7.47 per 1,000 views, about three times the $2.30 median⁵⁰. At $10 to $20 per 1,000 views, $10,000 a month needs 500,000 to 1 million views. At $2.30 it needs 4.3 million. Even in education and science, a quarter of channels are under $2.31⁸.
Viewers in countries where advertisers pay more⁹. Most of the broker-checked channels have viewers mainly in the US, UK, Canada and Australia, and often men over 45 to 55⁴⁶. One exception is a Spanish-language sports channel whose viewers are mainly in Mexico; it has the third-highest profit of the 20, $15,712 a month⁴⁶. The food channel reports 92% US viewers, mainly aged 55 and over, watching on smart TVs⁴⁸.
More of each view carrying an ad. AIR says its top 10% of earners came from gaming, music, entertainment and education alike. “What they had in common were the mechanics: more of their views showed an ad, each view carried more ads, their monetization settings were clean, and their audiences watched from higher-paying countries.”⁹ Channel size hardly mattered: the median was $2.14 per 1,000 views for channels with 10,000 to 100,000 subscribers and $2.04 for larger ones⁹. Longer videos have room for more ads: “A ten-minute video can hold multiple mid-rolls; an eight-minute video struggles to hold one.”⁸ AIR sells services to creators, and these are descriptions of its own data, not a controlled test.
Usually a small paid team, with the owner as editor. Most of the broker-checked channels pay freelancers: 17 of the 20 pay at least one, and where a number is given it runs from one to seven⁴⁶. Three are made by the owner alone; one of those owners reports 60 hours a week⁴⁶. The owner usually chooses topics and writes or checks scripts⁴⁶. Output varies widely, from about three videos a month to 40 to 80 a month; most of those that state it post one to five a week⁴⁶. Their views come from YouTube’s recommendations more than from search⁴⁶. Correction, 7 October 2026: an earlier version of this page said these channels run on 2 to 7 freelancers and post 2 to 7 videos a week. We re-read the 20 listings and those ranges did not hold. It also said cars appear six times; the count is seven vehicle channels.
Regular output at the start, and judgement after. For a channel under 1,000 subscribers, adding 52 or more videos in a year goes with a 30.77% chance of reaching 1,000, against 1.52% overall²⁹. Once a channel is established, more is not better. AIR compared the same 282 channels in busy and quiet months: “Entertainment channel posting 16+ videos a month earns 33% less total revenue than the same channel posting 1–2 videos a month, despite publishing 8 times more content.”⁵⁹ A small gaming channel’s pay per 1,000 views fell from $2.33 at 3 to 4 videos a month to $0.63 at 5 to 8⁵⁹. In a second AIR study of 20,000 channels, those with under 50 videos got more views per new upload than those with over 1,000 videos in 8 of 11 subjects⁶⁰. Among channels past 1,000 subscribers, vidIQ found median view growth of 2.18% a month for those uploading 12 or more times a month, against 0.53% for those uploading less than monthly⁵⁶. Even the fastest group’s typical growth is about 2% a month. All of these show links, not causes⁵⁶⁵⁹⁶⁰.
Experience or a head start. The operator in the Fortune case had about five years of building and selling channels, a partner who wrote the software, paid staff, and was early²³. The two-person narrated channel Not What You Think reached over 70 million views a month about a year after launch; one founder had been a full-time YouTuber since 2018, and they credit quality, consistency and Shorts⁶⁶. That profile is from 2022, before the AI wave, and gives no income figure⁶⁶.
Time without income. Ad income is zero until acceptance, and YouTube’s review takes about a month². Two of the broker-checked winners were first paid about three to five months after starting⁴⁷⁴⁹. Most channels take longer or never get there: 1.52% of small channels reach even 1,000 subscribers in 360 days, and 30.77% of those adding 52 or more videos²⁹. That is the subscriber part of the bar only.
Income beyond ads. The detailed accounts say most of the money came from elsewhere: a course in the 2022 report²², sponsors and products in the two-year account²⁴. One of the tweet authors writes that he would build a channel around a product of his own, not ads alone⁴².

The window may be closing for the templated route. The Fortune operator says copycats crowded his subject within days of his showing the method, and he expects individual creators to have until about 2027²³. The Hollywood Reporter says operators are hiring on-camera hosts because faceless channels are losing ad income⁵². Against that, vidIQ’s gaming study found that the most crowded game subjects had higher, not lower, shares of new channels reaching 1,000 subscribers⁵³. Whether crowding is rising has not been measured.

What it costs

You have to cover production the whole time, in one of two ways.

Do it yourself. Our estimate is $6 to about $190 a month in AI tools²⁰, or roughly $140 to $300 a month with research tools and stock footage added⁶³⁶⁴. Plus your hours. The two-year account reports over 2,000 hours²⁴. The writer above reports 4 to 6 hours a video⁶⁵.
Pay others. About $870 to $1,700 a month on our Fiverr estimate for short videos²¹. The two-year account reports $600 to $2,500 per video for longer ones²⁴. Most of the broker-checked channels pay one to seven freelancers; the profits above are after those costs⁴⁶. Their reported costs have a median of $731 a month and run from $17 (an owner working alone) to $3,144 (a team of seven)⁴⁶. Three listings give a cost per video: $45 to edit one, $60 to $80 for a complete 15-minute video, and about $309 for a researched long video⁴⁶.
Run a network. The $6,500 a month in the Fortune case is a staffed network of channels, not a newcomer’s cost²³. A product with the same name as the software Fortune names sells for $149 to $849 a month; we have not confirmed it is the same product⁶².

How to tell early which side you are on

These checks are our judgement, built from the figures above.

Has one of your first ten long videos passed 10,000 views? About three in ten committed new channels manage that⁵⁴. A channel with none is in the majority, and not yet in the group that pulls away.
Monthly views when you reach 1,000 subscribers. Under 5,000 a month is the common case²⁸. Over 20,000 puts you in the top 4.2%²⁸. A channel heading for 435,000 views a month needs single videos reaching tens of thousands.
Click-through rate. This is the share of people shown your video’s picture and title who click it. YouTube says: “Half of all channels and videos on YouTube have an impressions CTR that can range between 2% and 10%.”⁶¹ It says the range is wider for new channels and for videos under 100 views⁶¹. AIR’s warning signs for a failing video are a click-through rate under 0.5% and viewers watching under 35% of the video on average⁶⁰.
Watch hours against the bar. Only the last 12 months count². If three months of regular uploads give you under a quarter of the hours (1,000 of 4,000, or 2,000 of 8,000 from 1 February 2027⁴), you will not pass at that pace. Viewing has to grow.
Could a viewer tell your videos apart? If each one uses the same template, voice, music and AI script, it matches YouTube’s own example of what is not paid and may be removed¹⁷¹⁸.
Money out against views in. Track spending per 1,000 views. If it stays above about $2.30 once you are being paid, the channel loses money at the median rate⁸.
Whether the plan needs you to buy something first. The ad arithmetic above is free to check. No seller we found publishes what share of their students reach any income level.

What we did not verify

Four items that were on this list in the earlier version are now settled and have moved into the page: the three vidIQ studies and the random-sample paper were read and their figures confirmed (Step 4), TubeLab’s prices were read (“Who makes money from it”), and the Hollywood Reporter article was read in an archived copy.

How many people who try this succeed. No source measures views or earnings across people who attempted a faceless channel. We cannot give a percentage. The broker listings show winners only.
The income claims in the tweets. We did not identify or look up any channel behind “$2M+”, “$17,292 per month” or the others. We did not see the tweets’ images, videos or guides, and several tweet texts are cut off in our copy. Tweet dates are worked out from ID numbers, not read from the tweets.
Rates for faceless or AI channels specifically, and by viewer country. AIR’s data is for established channels of a services firm, its raw data is not public, and we did not find where its channels’ viewers live. AIR states what its top earners share but publishes no top-against-bottom numbers. A US rate of about $5 per 1,000 views appears on one site that does not disclose its method; we did not open it and do not use it. A US Shorts figure of about $0.33 was seen only in a search summary and is not used.
The vidIQ studies. They opened only through a reader service that fetches the page for us, not directly. vidIQ sells a YouTube tool. Its samples are its own index, not all of YouTube. None of its studies separates faceless or AI channels. Its dollar ranges at 1,000 subscribers are its estimates, not payouts. The page on pay at 1,000 subscribers shows a 2025 publish date with data labelled May 2026; we did not see when it was updated.
The broker listings. We did not see the records behind any Empire Flippers listing and did not identify any channel. That the broker checks revenue, traffic and expenses is its own statement, read on a listing page. “Sold” does not show the final price. The count of 20 faceless channels is our keyword match on listing summaries; at least one matched listing mentions a contractor presenting on camera. The count of 17 that pay freelancers is our reading of the listing summaries; one of the 17 says only that half the work is outsourced. Two of the three listings in the table were read from the broker’s listings data, not their listing pages. The broker’s marketplace page showed different monthly figures for some listings, perhaps for a shorter recent period; we used the averages in the listings data.
The Flippa listings. Both were read through a tool that summarises text; the celebrity and mystery listing was re-read through a reader service on 7 October 2026. Flippa’s “verified” label was not independently checked. Two other Flippa listings marked sold did not show a sale price or accounts to us.
The Fortune case. Fortune’s check was a reporter’s review of screenshots and payout records, not an audit. The costs and margins are the operator’s own statement. The name and price of the course he told Fortune he offers were not found, and we do not know whether it is paid.
The $250,000 a month in the Hollywood Reporter. The article does not say the reporter checked records, and we saw no other account of the copyright dispute or how it ended. We did not confirm whether the operator sells anything on the subject. The article was read in an Internet Archive copy and again on the live page on 7 October 2026.
A second checked success. We found no second success case checked by a journalist. Business Insider and the Guardian could not be opened by our tools. Widely repeated earnings for well-known faceless channels are third-party estimates from view counts and are not used.
What successful documentary channels spend per video. We found no first-hand account beyond the broker listings: their team sizes, their reported monthly costs and three stated costs per video. Those are the sellers’ figures as the broker lists them.
Whether crowding is rising. We have one undated figure from YouTube and one operator’s account. We found no measured trend for faceless or AI channels.
Six-figure channels. The rise of more than 7% in US channels earning six figures was seen only in a trade report. YouTube’s own post on the report does not contain it, and we did not open the report.
Reddit. Our tools could not open it, so no creator forum threads were read. This is the largest gap in the accounts from people who tried.
Exact quotes. Many pages were read through a tool that extracts or summarises text. Quotes from YouTube’s three policy pages were read as full text. Quotes from Fortune, AIR and the random-sample paper were checked against the page text. Other quotes, including those from the vidIQ, TubeLab, TubeGen, Storyblocks, Tubefilter and YouTube click-through pages, were read through such a tool and were not checked word for word against the page text.
Refused pages. Fiverr, Upwork, nytimes.com, Goldman Sachs and the sign-up page for the guide refused us. vidIQ refused direct reading. The New York Times piece was read in a republished copy. Freelancer rates were not re-priced.
Prices. invideo’s page showed different prices on different visits (it may vary by country or offer). Storyblocks showed euro prices that may vary by country, and Epidemic Sound’s prices did not load. The vidIQ, TubeLab and TubeGen prices were read through a summarising tool. The regular price of any “usually $87” guide, the cost of 30 to 50 minute videos and done-for-you agency prices were not captured. The euro rate of $1.15 is a round figure, not looked up.
What two named accounts sell. We found no paid course by the newsletter’s author and could not read his shop page; we read only the first tweet of his thread. A product with the same name as the software in the Fortune case is sold to the public, but its website does not name the people in the article, so the link is not confirmed.
Reports we left out or could not confirm. A widely repeated case of a creator spending $26,311 to earn $15,218 is attributed by one blog to a named creator, but the page it cites no longer exists and we found no original. A figure of about 12 million channels terminated in 2025 for all kinds of spam was seen only second-hand; Google’s Transparency Report did not load for our tools. Course-sales and refund figures for two other sellers were seen only in a third-party summary and are not used.
Removed channels’ income. Revenue figures for AI channels in Kapwing’s report are estimates from view counts, not payouts³⁰. YouTube publishes no count of AI or faceless channels removed or refused.
Rules after 1 February 2027. We read the new thresholds but not what else the new terms change. YouTube has not published the details of the promised bonuses.
Regulators outside US federal sources. State, UK and EU authorities and private lawsuits were not searched.
Tax and payment thresholds, which change what a channel keeps.
Whether the people named have seen this page. The people and companies named here were not contacted before publication.

Sources

“Read at source” means we opened the page itself on 7 October 2026. “Secondary” means we read an article or summary about the source, not the source itself. “Estimate” means our own sum. “Through a reader service” means a service fetched the page for us because the site refused our tools directly.

Does It Pay collection of 146 tweets on this scheme, collected 2026-10-06; individual tweets linked above. Read from our collected copy on 2026-10-07, not from the live tweets. Counts by label are estimates.
YouTube Help, “YouTube Partner Program overview and eligibility”. https://support.google.com/youtube/answer/72851?hl=en . Undated page. Read at source on 2026-10-07.
YouTube Help, expanded Partner Program tier. https://support.google.com/youtube/answer/13429240?hl=en . Undated page. Read at source on 2026-10-07.
YouTube Blog, Partner Program updates for 2027. https://blog.youtube/news-and-events/youtube-partner-program-updates-2027-new-opportunities-earn/ . 2026-08-10. Read at source on 2026-10-07. Also confirmed at https://support.google.com/youtube/answer/12843009 .
YouTube Help, revenue share. https://support.google.com/youtube/answer/72902?hl=en . Undated page. Read at source on 2026-10-07.
YouTube Help, Shorts ad revenue sharing. https://support.google.com/youtube/answer/12504220?hl=en . Undated page. Read at source on 2026-10-07.
YouTube Help, RPM and CPM. https://support.google.com/youtube/answer/9314357?hl=en . Undated page. Read at source on 2026-10-07.
AIR Media-Tech, “How much does YouTube really pay in 2026? Real RPM data from 300 channels”. https://air.io/en/air-data-findings/how-much-does-youtube-really-pay-in-2026-real-rpm-data-from-300-channels . 2026-07-01. Read at source on 2026-10-07; the overall median, the November and January medians and the subscriber range were re-read in the full article the same day. Table of groups: “What is YouTube RPM”, https://air.io/en/youtube-glossary/what-is-youtube-rpm-full-breakdown , updated 2026-08-25, read at source on 2026-10-07.
AIR Media-Tech, “YouTube monetization statistics 2026”. https://air.io/en/air-data-findings/youtube-monetization-statistics-2026 . 2026-10-01. Read at source on 2026-10-07.
Business Insider, republished by Yahoo. https://ca.news.yahoo.com/much-money-youtube-pays-1-120000576.html . 2023-09-14. Secondary (creators’ own reports to a journalist).
vidIQ, “How much does YouTube pay per view”. https://vidiq.com/blog/post/how-much-does-youtube-pay-per-view/ . 2026-04-17. Secondary; no method stated.
McGrady, Zheng, Curran, Baumgartner and Zuckerman, “Dialing for Videos”, Journal of Quantitative Description: Digital Media 3. https://journalqd.org/article/view/4066 . 2023-12-20. The study behind¹³. Read at source on 2026-10-07 (the PDF from the journal, https://journalqd.org/article/download/4066/3766 ); an earlier version of this page could not open it.
Ethan Zuckerman, “How big is YouTube?”. https://ethanzuckerman.com/2023/12/22/how-big-is-youtube/ . 2023-12-22. Read at source on 2026-10-07.
vidIQ, “YouTube subscriber benchmarks 2026”. https://vidiq.com/research/youtube-subscriber-benchmarks-2026/ . 2026-07-20. Read at source on 2026-10-07, through a reader service; direct reading was refused. Earlier read second-hand.
YouTube Blog, chief executive’s letter for 2026. https://blog.youtube/inside-youtube/the-future-of-youtube-2026/ . 2026-01-21. Read at source on 2026-10-07. The $700 a month figure is our estimate from this source and⁴.
YouTube Blog, Partner Program in Armenia. https://blog.youtube/news-and-events/youtube-partner-program-armenia/ . 2026-05-11. Read at source on 2026-10-07.
YouTube Help, “YouTube channel monetization policies”. https://support.google.com/youtube/answer/1311392?hl=en . Update note dated 2025-07-15; page otherwise undated. Read at source on 2026-10-07.
YouTube Help, “Spam, deceptive practices, and scams policies”. https://support.google.com/youtube/answer/2801973?hl=en . Undated page. Read at source on 2026-10-07.
YouTube Help, disclosing altered or synthetic content. https://support.google.com/youtube/answer/14328491?hl=en . Undated page. Read at source on 2026-10-07.
Price pages: https://elevenlabs.io/pricing , https://invideo.io/pricing/ , https://www.heygen.com/pricing . Undated pages. Read at source on 2026-10-07. The combined tool costs are our estimate.
Fiverr listing for faceless videos. https://fiverr.com/taimur_1214/create-faceless-youtube-videos-with-script-writing-editing-and-voiceover . Undated. Secondary: search summary on 2026-10-07; the page refused us. Monthly totals are our estimate.
New York Times article, republished by Jordan News as “You too can strike it rich on YouTube”. https://www.jordannews.jo/Section-129/Technology/You-too-can-strike-it-rich-on-YouTube-20298 . 2022-08-10. Secondary (republished copy, read on 2026-10-07).
Fortune, on Adavia Davis’s faceless channels. https://fortune.com/2025/12/30/ai-slop-faceless-youtube-accounts-adavia-davis-user-generated-content/ . 2025-12-30. Read at source on 2026-10-07.
Third-party summary of the video “I Tried Faceless YouTube Automation for 2 Years to Find Out if the Gurus Were Lying” (youtube.com/watch?v=jUTOst0tb7Q). https://lilys.ai/ja/notes/openai-agent-builder-20260208/faceless-youtube-automation-2-year-review . Video dated 2025-09-14. Secondary.
Third-party summary of an earlier video by the same creator (youtube.com/watch?v=cKTU4fcttZ0). https://lilys.ai/de/notes/youtube-automation-20251207/truth-youtube-automation-results . Video dated 2024-07-16. Secondary.
Third-party summary of the video “The TRUTH About YouTube Automation” (youtube.com/watch?v=W2Ur0K817ks). https://lilys.ai/en/notes/openai-agent-builder-20260208/youtube-automation-truth . Video dated 2026-01-23. Secondary.
“My YouTube Automation Journey pt. 2”, Vocal Media. https://vocal.media/motivation/my-you-tube-automation-journey-pt-2 . 2022 (page shows only “published 4 years ago”). Read at source on 2026-10-07.
vidIQ, on pay at 1,000 subscribers. https://vidiq.com/blog/post/youtube-pay-1000-subscribers/ . Published 2025-10-23, with data labelled May 2026; update date not seen. Read at source on 2026-10-07, through a reader service; direct reading was refused. Earlier read second-hand. Its dollar figures are vidIQ’s estimates.
vidIQ, uploads and reaching 1,000 subscribers. https://vidiq.com/research/youtube-uploads-1000-subscribers/ . 2026-09-11. Read at source on 2026-10-07, through a reader service; direct reading was refused. Earlier read second-hand.
Kapwing, report on AI-generated videos in trending channels. https://www.kapwing.com/blog/ai-slop-report-the-global-rise-of-low-quality-ai-videos/ . 2025-11-28. Read at source on 2026-10-07. Its revenue figures are third-party estimates.
TheWrap, on the termination of Screen Culture and KH Studio. https://www.thewrap.com/industry-news/tech/youtube-terminates-screen-culture-kh-studio-over-ai-fake-movie-trailers/ . 2025-12-18. News report read on 2026-10-07; secondary for the YouTube statement.
Net Influencer, on removed AI channels. https://www.netinfluencer.com/youtube-removes-major-ai-slop-channels-with-combined-9-8m-in-estimated-annual-revenue/ . 2026-01-29. News report read on 2026-10-07; secondary for Kapwing’s count.
Tubefilter, on removed AI channels. https://tubefilter.com/2026/01/29/youtube-ai-slop-channel-crackdown-bans/ . 2026-01-29. News report read on 2026-10-07; secondary for Kapwing’s count.
Net Influencer, on YouTube’s trust and safety head explaining the rules. https://www.netinfluencer.com/youtube-trust-and-safety-chief-details-three-categories-of-content-barred-from-monetization/ . 2026-07-21. Secondary (reports a YouTube video we did not watch).
PPC Land, on YouTube’s July 2025 clarification. https://ppc.land/google-clarifies-youtube-monetization-policies-amid-creator-confusion/ . 2025-07-11. Secondary (quotes a YouTube forum reply that did not load for us).
PPC Land, on YouTube’s review of terminations. https://ppc.land/youtube-addresses-creator-concerns-on-content-moderation-and-appeals/ . 2025-11-13. Secondary.
Empire Flippers, “What to look for when buying a faceless YouTube channel”. https://empireflippers.com/what-to-look-for-when-buying-a-faceless-youtube-channel/ . Updated 2026-08-19. Read at source on 2026-10-07.
Flippa listing 12813083, faceless documentary channel. https://flippa.com/12813083-faceless-documentary-youtube-channel-with-11m-views-235k-revenue-98-profit-margin-and-20-rpm-in-a-premium-history-niche . Active listing, undated. Read at source on 2026-10-07.
Andrew Bolis’s website. https://andrewbolis.com/ . Undated page. Read at source on 2026-10-07.
WannerCashCow pages. https://wannercashcow.com , https://www.skool.com/wannercashcow , https://wannercashcow.com/refund-policy (effective 2026-03-05). Read at source on 2026-10-07.
Trustpilot, Automate Channels. https://ca.trustpilot.com/review/automatechannels.com . Reviews dated November 2022 to April 2026. Read at source on 2026-10-07.
Adam Del Duca newsletter, “The $10,000/Month Faceless YouTube Channel System”. https://adamdelduca.beehiiv.com/p/the-10-000-month-faceless-youtube-channel-system . 2026-06-10. Read at source on 2026-10-07.
ElevenLabs affiliate page. https://elevenlabs.io/affiliates . Undated page. Read at source on 2026-10-07.
Federal Trade Commission, complaint in FTC v. Growth Cave (C.D. Cal., 2:25-cv-01115). https://ftc.gov/system/files/ftc_gov/pdf/GrowthCave-Complaint.pdf . 2025-02-10. Read at source on 2026-10-07.
Federal Trade Commission press releases on Growth Cave. https://www.ftc.gov/news-events/news/press-releases/2025/03/ftc-takes-action-stop-sprawling-growth-cave-business-opportunity-credit-repair-scam (2025-03-07) and https://www.ftc.gov/news-events/news/press-releases/2026/01/ftc-secures-settlement-banning-growth-cave-defendants-marketing-selling-business-opportunities (2026-01-27). Read at source on 2026-10-07.
Empire Flippers, public listings data for businesses with YouTube income. https://api.empireflippers.com/api/v1/listings/list?monetizations=YouTube&limit=100 ; public page https://empireflippers.com/marketplace/?monetization=youtube . Live data, undated; listings first listed June 2024 to September 2026. Read at source on 2026-10-07. The count of faceless channels and the medians are our own sums from this data.
Empire Flippers listing 83979, faceless niche-news channel. https://empireflippers.com/listing/83979/ . First listed 2025-03-24. Read at source on 2026-10-07 (from the listings data in⁴⁶).
Empire Flippers listing 93968, faceless food channel. https://empireflippers.com/listing/93968/ . First listed 2026-05-04. Read at source on 2026-10-07.
Empire Flippers listing 88223, faceless automotive channel. https://empireflippers.com/listing/88223/ . First listed 2025-09-19. Read at source on 2026-10-07 (from the listings data in⁴⁶).
Empire Flippers listing 93478, faceless trucking channel. https://empireflippers.com/listing/93478/ . First listed 2026-04-22. Read at source on 2026-10-07 (from the listings data in⁴⁶).
Flippa listing 12650859, faceless celebrity and mystery channel. https://flippa.com/12650859-high-profit-faceless-youtube-channel-in-celebrity-mystery-958k-revenue-rpm-10-usd-110m-views-us-audience-team-included . Active listing, undated. Read at source on 2026-10-07, through a summarising tool.
Andrew Zucker, The Hollywood Reporter, on faceless creators and AI. https://www.hollywoodreporter.com/business/digital/faceless-creators-youtube-ai-damage-1236617586/ . 2026-06-13 (10 June print issue). Read on 2026-10-07 in the Internet Archive copy, https://web.archive.org/web/2026/https://www.hollywoodreporter.com/business/digital/faceless-creators-youtube-ai-damage-1236617586/ ; the live page refused some of our tools. Read again on the live page on 2026-10-07.
vidIQ, Minecraft, Roblox and small gaming channels. https://vidiq.com/research/minecraft-roblox-small-creators/ . 2026-09-15, updated 2026-09-22. Read at source on 2026-10-07, through a reader service.
vidIQ, study of new channels’ first ten videos. https://vidiq.com/research/youtube-first-10-videos-study/ . 2026-07-28, updated 2026-07-30. Read at source on 2026-10-07, through a reader service.
vidIQ, first-week views of long videos. https://vidiq.com/research/youtube-first-week-views/ . 2026-09-16. Read at source on 2026-10-07, through a reader service. Dollar figures drawn from it are our estimate.
vidIQ, upload frequency study. https://vidiq.com/research/youtube-upload-frequency-study/ . 2026-07-08. Read at source on 2026-10-07, through a reader service.
PPC Land, on YouTube’s 2025 US impact report. https://ppc.land/youtube-gains-50-000-us-jobs-to-hit-540-000-in-2025-report/ . 2026-07-16. Secondary: the report itself was not opened, and YouTube’s blog post about it does not contain the figure.
YouTube, press page. https://blog.youtube/press/ . Undated page. Read at source on 2026-10-07.
AIR Media-Tech, “Does posting more on YouTube hurt your RPM”. https://air.io/en/air-data-findings/does-posting-more-on-youtube-hurt-your-rpm . 2026-07-14. Read at source on 2026-10-07.
AIR Media-Tech, “More videos don’t mean more views: we analyzed 20,000 YouTube channels”. https://air.io/en/audience-growth/more-videos-dont-mean-more-views-we-analyzed-20000-youtube-channels-to-find-that-out . 2026-06-17. Read at source on 2026-10-07.
YouTube Help, impressions and click-through rate. https://support.google.com/youtube/answer/7628154?hl=en . Undated page. Read at source on 2026-10-07.
TubeGen AI website. https://tubegen.ai/ . Undated page. Read at source on 2026-10-07, through a summarising tool.
TubeLab pricing page. https://tubelab.net/pricing . Undated page. Read at source on 2026-10-07, through a reader service.
Price pages: https://vidiq.com/pricing/ (through a reader service) and https://www.storyblocks.com/pricing . Undated pages. Read at source on 2026-10-07. The combined cost is our estimate.
“My personal experience with YouTube automation channels and how much money I made”, Drew’s Review. https://drews-review.com/my-personal-experience-with-youtube-automation-channels-and-how-much-money-i-made/ . 2026-04-30. Read at source on 2026-10-07. The author’s own statement; he sells a course.
Tubefilter, “Creators on the rise: Not What You Think”. https://www.tubefilter.com/2022/08/03/creators-on-the-rise-not-what-you-think/ . 2022-08-03. Read at source on 2026-10-07.

Corrections

If you are quoted or named on this page and think something is wrong, want your reply shown beside it, or want to be removed, write to [email protected]. We aim to reply within 14 days, and always within one month. Factual errors are corrected with a dated note here. A sentence that is seriously disputed comes down while we check it.

Change log:

2026-10-07: Replaced “When it can work” with “The upside” and “What it takes to compete”. Added broker-checked channels listed for sale and other new evidence (sources 46 to 66). The three vidIQ studies and the random-sample paper, earlier read second-hand, were read at source and their figures confirmed; the overall 1.52% figure was restored; the wording “weekly or more often” was removed. TubeLab’s prices and the Hollywood Reporter article were read. Verdict and label changed from “Unlikely for a newcomer” to say who the scheme pays.
2026-10-07 (review): Corrected the description of the 20 broker-checked channels after re-reading the listings: 17 of 20 pay freelancers (one to seven) and three owners work alone, not “2 to 7 freelancers” for all; output runs from about three videos a month to 40 to 80, not “2 to 7 a week”; seven vehicle channels, not six. Added their reported monthly costs. Corrected the second Flippa listing: $913,731 is revenue since the channel began, and about $895,000 is the past year. Reworded the Fortune case to say the reporter reviewed screenshots, and that the 2 million views a day is the operator’s statement. Verdict and label reworded.
2026-10-07: wording about named accounts and companies reviewed; inferences about named accounts and companies were replaced with what was seen, one small account and one operator’s name were removed, and the Hollywood Reporter, Fortune and refund-policy pages were read again.