Elevate is a community dedicated to making money through faceless YouTube.
New York
Joined August 2024
- Tweets1.2K
- Following3
- Followers1.9K
- Likes1.5K
Your channel is stuck between two kinds of content, and you haven't picked a lane.
Evergreen: topics people will still search for next year.
Timely: topics people want this week.
Beginners mix both with no plan, then can't read their own numbers.
Evergreen starts slow and keeps paying. In my experience, it builds the floor under a channel.
Timely spikes fast and fades fast. It can get you noticed, but you start from zero every week.
Mistake 1: All timely.
Every video is a coin flip with an expiry date.
The fix: make timely the minority. Our rule of thumb is 70% evergreen, 30% timely. That's our split, not YouTube's.
Mistake 2: All evergreen, no demand testing.
The fix: use one timely video per batch to test a new angle. If it pulls, build an evergreen version.
Mistake 3: Judging both on the same clock.
The fix: read timely videos in days and evergreen videos in weeks. Don't trust retention on either until 3,000-5,000 views.
The order:
Anchor on evergreen. Use timely as your test lane. Read each on its own clock.
Timely gets the spike.
Evergreen builds the channel
Repackage a flop in 30 minutes:
Check the gate first. Impressions fine but clicks weak is a packaging problem. Low impressions or weak retention are NOT. Don't repackage those.
Write 5 new titles that match what the video actually delivers.
Change ONE thing: title or thumbnail, not both.
Save the old version and the date in your log.
Wait 7 days. Compare impressions and CTR against the week before.
Repackage the packaging.
Never change the promise.
Most creators can't tell which YouTube notice they just got.
So they panic over the harmless one and shrug at the dangerous one.
Three notices look almost identical in your inbox. Only two can hurt the channel.
Notice 1, the Content ID claim.
An automated match against someone's registered work. It affects ONE video. The rightsholder can take the revenue, track the views, or block it.
It is not a strike.
The fix: read what the claim actually says. Swap the track or cut the clip if you want the video clean. Only dispute it with a real basis, because a bad dispute can escalate into a formal takedown request.
Notice 2, the copyright strike.
This follows a valid copyright removal request. The video comes down and the strike sits on the channel.
Three active copyright strikes put the channel at risk of termination.
The fix: find the exact asset and stop using it anywhere else on the channel.
Notice 3, the Community Guidelines warning or strike.
Separate system, separate counter. Spam, misleading metadata, harassment, graphic content. A first offense often starts as a warning. A repeat inside the window becomes a strike, and 3 strikes in 90 days can end the channel.
Deleting the flagged video does NOT erase the warning or strike.
The fix: appeal if it's wrong. If it's right, fix the habit that caused it.
The order matters.
Open Studio. Identify the notice type FIRST. Respond to that one. Screenshot it with the date and log which asset caused it.
Most of this traces back to one habit: unlicensed music and clips. Licensed tracks and YouTube's Audio Library come with receipts.
I'm not a lawyer, and fair use is decided case by case. When in doubt, don't use it.
Panic costs views.
Triage costs 10 minutes.
You're fixing your videos without knowing where your views come from…
That's treating a symptom you haven't found.
Open YouTube Studio. Analytics. Reach. Traffic source types.
That one card tells you what kind of channel you're running.
Browse: YouTube is pushing you on its own.
Heavy browse means your packaging works. It also means you live on the algorithm's mood.
The fix: change one thing at a time so you know what moved it.
Suggested: you're showing up next to other videos.
The fix: make topics adjacent to proven videos in your niche. Same viewer, new angle.
Search: someone typed it in.
In my experience, search is the slowest source and the most durable. It builds a floor under the channel.
The fix: use the exact phrase a person would search in the title, then confirm it in the first 30 seconds.
External: someone sent them.
The fix: collabs and communities live here.
Rule of thumb: if one source brings 80%+ of your views, you're fragile. That's our number, not YouTube's.
The order:
Read the card. Find the dominant source. Build a second one.
One source is a spike.
Two sources is a channel.
YouTube is doubling the watch hour bar on February 1, and a lot of people are treating it like a death sentence.
It isn't. But plenty will quit over it without ever running the numbers, so let's run them.
The facts first. From February 1, 2027, new applicants need 8,000 watch hours in the past 365 days, up from 4,000. The Shorts route doubles to 20 million views in 90 days. The 1,000 subscriber requirement stays. Existing Partner Program members are NOT affected.
8,000 hours is 480,000 minutes. That's about 22 hours of watch time every single day for a year.
Here's what that looks like if every view looked the same:
An 8 minute video at 45% retention gives you 3.6 minutes per view. You need about 133,000 views.
A 20 minute video at 40% gives you 8 minutes per view. About 60,000 views.
A 40 minute video at 35% gives you 14 minutes per view. About 34,000 views.
A 90 minute documentary at 30% gives you 27 minutes per view. About 17,800 views.
Same bar. Four completely different jobs.
Those retention numbers are illustrations, not benchmarks. Plug in your own.
Now the part most people miss.
Long videos are the cheapest road to hours, but only if they hold. That same 90 minute video at 8% retention gives you 7.2 minutes per view. You're back to about 67,000 views.
A padded 90 minutes loses to a tight 40 every time.
And the clock doesn't reset on February 1. Hours you earn now still count toward the trailing 365 days.
So test the channel first. Wait 72 hours. Keep the 2 to 3 that get impressions. Then build the format your math supports.
Hours aren't a volume problem.
They're a retention problem.
Your documentary can be 100% accurate and still put your channel at risk.
Accurate and careful are not the same thing.
Beginners fact-check the script, then publish a video that names a real person, implies guilt, and rests on one source.
Mistake 1: Stating allegations as facts.
"He did it." "He was charged with it." "Prosecutors alleged it."
Those are three different claims.
The fix: match your wording to your source. Convicted, charged, accused, and rumored are four different words.
Mistake 2: One source per serious claim.
The fix: two independent sources minimum, and at least one primary where possible. That's our rule of thumb.
Mistake 3: Letting the visuals say what the script doesn't.
A real person's photo over a line about a crime is an accusation.
The fix: use neutral footage whenever the script discusses wrongdoing.
Mistake 4: No paper trail.
The fix: keep a log. Every claim, its source, the date you checked it. If anyone disputes it, you have the receipt.
The order:
Source it. Word it precisely. Check what the visuals imply. Log it.
I'm not a lawyer, and rules vary by country. For anything serious involving a living person, get real legal advice.
Accurate gets you views.
Careful keeps the channel.
Elevate retweeted
14 dead channels taught me everything I know about faceless YouTube.
Every single one was a different error I did not understand at the time.
And 99% of the people in my DMs right now are making the same 9 errors I made 18 months ago. Here is the full list with the fix for each, so you find yours before burning another 3 months.
1. Skipping the impression test. I uploaded 15 produced videos into a channel the algorithm wrote off on day 1. Fix: film the floor for 3 minutes, title it "a walk in the park", wait 72 hours. 300+ impressions means the channel works. Under 250, walk away.
2. Chasing untapped niches. Zero competition means nobody figured out how to make the niche work. Fix: find 10 channels doing 1,000+ views per video in the same angle. Under 5, leave.
3. A dirty proxy. High fraud score IPs kill your reach silently, no email, no flag. Fix: run browserleaks from inside the profile before uploading anything. 30 seconds.
4. Batch uploading a new channel. 5 videos in 3 days reads identical to a spam farm. Fix: 1 video every 48 to 72 hours for the first 2 weeks, then ramp.
5. Verifying advanced features with ID. You are tying your face to a portfolio. When Google terminates, they terminate YOU rather than the channel. Fix: skip it and let the features activate after monetization.
6. Generic prompts instead of JSON templates. "Write me a script about world war 2" produces the same script as 10,000 other channels this week.
7. Picking a voice you personally like. Your 55+ romance audience wants a deep male narrator. Your preference is irrelevant.
8. Mixing niches on 1 channel. By video 10 the algorithm has conflicting data on 3 audiences and serves you to none of them.
9. Quitting at video 8. The audience model does not mature until video 12 to 15 on a healthy channel.
Fix the infrastructure before touching the content.
9 out of 10 channels I audit with 0 views have an infrastructure problem, not a content problem.
Full system behind every fix, plus the niche finder and the templates, sits inside elevate(.uno)
Or DM me "FIX" and I'll tell you which of the 9 is killing your channel
Shorts will NOT fix your new long form channel.
In my experience, they bring a different viewer.
Someone swipes through a 30 second clip, watches it, and swipes on. They didn't choose you. The feed chose for them.
So a Short can hit 200k views while your long videos sit at 300.
That's not a growth engine. That's a second channel with a different audience.
Shorts and long form are also monetized differently, so check your own revenue report before you assume they pay the same.
The fix: pick your lane.
If your goal is watch time and long form revenue, build the long-form channel first. Treat Shorts as a side experiment, not the plan.
Views feel good.
The right views pay.
A channel with 1M views can earn less than a channel with 100k.
Most beginners track views and nothing else.
Then they get monetized and can't work out why the numbers don't match the effort.
Income is roughly views × RPM. RPM is revenue per 1,000 views. It swings with niche, audience country, video format, and season.
Illustrative math, not real data:
1M views at a $1 RPM = $1,000.
100k views at a $10 RPM = $1,000.
Same money. Ten times the audience.
Mistake 1: Choosing a niche on views alone.
The fix: treat RPM as a hypothesis before monetization. Test two angles in adjacent niches, then compare your own numbers once the revenue report unlocks.
Mistake 2: Reading the channel average and ignoring video level RPM.
The fix: sort your videos by RPM, not just views. Make more of the topics that pay.
Mistake 3: Padding length for ads.
Videos over 8 minutes can carry mid-roll ads. But a padded 12 loses to a tight 6 when retention collapses.
The fix: add length only when the story needs it.
The order:
Get views. Read RPM. Double down on what pays. Cut what doesn't.
Views get you noticed.
RPM pays the bills.
already shown impressions:
Hour 1: pick 4 topics, write 3 title options each.
Hours 2-3: write 4 scripts from your template.
Hour 4: record or generate all 4 voiceovers.
Hour 5: visuals and edits, same style every time.
Hour 6: thumbnails, descriptions, schedule one upload every 48 to 72 hours.
The hours are a rough template. Adjust them to your format.
Don't batch for an untested channel. Test first.
Elevate retweeted
YouTube DOUBLED the monetization requirement to 8,000 watch hours and everyone is acting like it’s nearly impossible to hit.
It isn't. But plenty of people will quit over it without ever running the numbers, and I'm not going to pretend that's bad for me. Let’s run the numbers.
8,000 hours is 480,000 minutes of watch time.
A 10 minute video at 50% retention hands you 5 minutes per view. You need 96,000 views inside 12 months. On a brand new channel with no trust and no subscribers, good luck.
A 45 minute video at 40% retention hands you 18 minutes per view. Now you need around 26,600.
A 3 hour sleep video at 20% retention hands you 36 minutes per view. The target drops to roughly 13,300.
An 8 hour sleep video where the viewer presses play at bedtime and leaves the room hands you a FULL NIGHT from one person.
Somewhere around 1,000 to 2,000 overnight plays clears the entire threshold.
Here are 3 niches you can go into where this is easiest:
- AI stories repositioned as sleep content, where Sherlock compilations pull 400k+ views and a 1,300 subscriber anime lore channel hit a 100x outlier on one 2 hour video
- Sleep music, where delta wave frequencies between 0.5 and 4Hz are doing 60M views on a single 5 hour upload
- Educational content framed as a sleep aid, where a video titled "Boring Psychology For Sleep" is at 1.9M, because boring is the SELLING POINT for this audience
One honest warning before you head into the music niche.
Music solves watch hours almost violently but barely moves subscribers, since nobody subscribes to the channels they fall asleep to.
Lead with stories or educational, where a viewer forms a relationship with the channel, then stack music underneath once subs are handled.
The validated pockets and reference channels are inside the Niche Finder at elevate(.uno)
Or DM me "SLEEP" and I'll tell you which of the 3 fits your setup
Watching 100 videos in your niche before making one is not research…
It's hiding with a notebook.
I see beginners spend a month studying competitors. Screenshots, spreadsheets, thumbnail swipe files. ZERO uploads.
Watching a video tells you what worked for someone else's channel, audience, and timing. It tells you NOTHING about yours.
Your own data is the only research that counts.
Watch 5 or 10 videos in the niche. Not 100. Note the structure, the pacing, the packaging.
Then publish.
Then read your impressions and your retention graph.
One real upload teaches you more than 30 hours of studying other people's.
Research is something you do between uploads. Not instead of them.
Study less. Ship sooner. Let your own numbers argue back.
I have never picked a niche because I was passionate about it, and I've never regretted it.
The ritual everybody performs is the least reliable method you could possibly use.
Write down your interests. Pick the one you love most. Spend a month building a channel around it.
Nothing in that process asks the one question that matters: do viewers already want this?
Here's what I do instead. It costs about 30 minutes.
Find small channels in your topic area. Under 1,000 subscribers.
Then look for any video on those channels with 50,000+ views.
Those are outliers. Collect 20 of them. (Those numbers are our rule of thumb, not an official line.)
The logic underneath is what makes this work.
A channel with 400 subscribers has no audience that explains 50,000 views. No fanbase, no brand, no history.
So the TOPIC did the work. Or the format. Or the angle.
That's what you borrow. The logic, never the script.
Now the part that turns this from a trick into a system.
Sort the 20 outliers into clusters. Same topic type, same structure, same packaging.
A cluster that shows up 5+ times is a niche candidate. Something that shows up once is probably luck.
Then test it before you commit. Upload a simple video, wait 72 hours, read impressions.
Sequence matters. Demand first. Channel second. Production last.
I see people spend 6 weeks building a niche they love that nobody is searching for, then blame the algorithm.
The niche wasn't wrong for them. It was wrong for viewers.
Passion is nice. Demand pays.
"Passive income" is the lie that kills most faceless channels…
The setup is front loaded. The work isn't.
You still pick the niche, test the channel, write the scripts, read the analytics, fix the intro, and upload on a rhythm.
The operators in our community making $10k $50k a month aren't passive. They run a system every single week.
The beginner mistake is expecting the channel to run itself after video 10.
Faceless means no face. It does NOT mean no work.
The fix: build a weekly routine.
Topic research. Script. Publish. Review the data. Change one thing.
Passive might come later.
Systems come first.
Elevate retweeted
Every AI story channel doing a million views a month is a blueprint sitting in public.
Find 10 channels in the same niche and angle (the same kind of story), each averaging 1,000+ views per video. Under 5, walk away.
Step 0: set-up
"Use NexLev MCP to find 10 YouTube channels in the [NICHE] space launched in the last 12 months with average views above 1,000 per video. Return the channel name, subscriber count, total videos, average views per video, estimated RPM range, primary audience demographic, and the titles of the top 3 outlier videos for each channel. Sort by average views descending."
Step 1: reverse-engineer the script.
Pull the transcripts from the target's 3 best videos (YouTube shows them under the video: "Show transcript") and paste them into Claude with:
"Analyze these 3 reference scripts from a top-performing channel in my niche. Study their structure, pacing, chapter length distribution, hook format, character introduction patterns, emotional beats, cliffhanger placement, and resolution type. Generate a JSON template defining total parts, words per part, hook type, cliffhanger placement, character registry format, setting parameters, voice directives, and resolution type."
Claude gives you a JSON template: a structured file that tells it how every script should be built. From then on, paste the JSON and a title into Claude and ask for the script. No other prompt needed.
People skip this step, type "write me a 40-minute story," and the model changes the characters by chapter 6.
Step 2: copy the production style.
Match the runtime to what the target runs. Match the voice: deep male for cowboy, female narrator for duke romance, documentary cadence for WW2.
For the thumbnail, attach one of the target's thumbnails in Claude and paste this. Then run the prompt Claude writes in an image tool like Nano Banana Pro or ChatGPT:
"I need a thumbnail image prompt for a YouTube video titled '[VIDEO TITLE]' in the [NICHE] niche. The thumbnail should 1) feature [NUMBER] characters with specific, exaggerated facial expressions conveying [EMOTION], 2) be set in [SETTING] with [LIGHTING CONDITIONS], 3) leave the top corners relatively uncluttered for title text overlay, 4) use [COLOR TEMPERATURE] tones matching the mood. Style reference: hyperrealistic cinematic film still, 16:9 horizontal, [GENRE DESCRIPTOR]. Include 8k detail, photorealistic skin texture, volumetric light, and shallow depth of field."
Step 3: build the channel.
Upload a boring 3-minute video titled "a walk in the park." Wait 72 hours. If YouTube showed it 300+ times from the homepage or suggested videos (check YouTube Studio), the channel works. Under 250, start a new one.
Then brand it: a unique name, a square profile picture, and a description rewritten from a competitor's:
"Here is a channel description from a competitor in my niche. Rewrite it for my new channel, [CHANNEL NAME], in the same tone, without copying any sentences."
Step 4: produce and upload on a loop.
Paste 10 of the target's titles into Claude:
"Based on the following titles, please generate 30 new titles following the same style, structure and keywords."
Script each title with your JSON, voice it, add the visuals and upload. 1 upload every 48-72 hours for 2 weeks, then ramp to 2-3 a day once YouTube starts pushing your videos.
Run the 5 steps on one channel this week, and you'll learn more than a month of guessing would teach you.
If you want me to pick the target channel with you, DM me "CLONE"
Elevate retweeted
This channel about bridges, dams and walls is doing ~3.7 million views a month.
It focuses on engineering, arguably one of the most BORING topics on planet Earth.
One video did 6.4 MILLION views, and the title worked because of 5 words you'd never think to write…
The channel is NextGen Manufacturing.
Its first video went up on December 14th 2025, and less than 10 months later it has 169k subscribers.
Its biggest video is called "How Japan Built the Impossible 400 km Great Wall that You've Never Heard Of." It has 6.4M views.
The 5 words are the end of that title: "that You've Never Heard Of."
Without those 5 words, it's a video about a wall in Japan. Easy to scroll past. With them, the viewer learns there's a 400 km wall he's never heard of, and he wants to know how he missed it.
The channel starts a lot of its titles the same way: "How [a country] Built the Impossible…". Its Dubai video, "How Dubai Built the $1B Impossible Burj Al Arab on Open Water", did 2.6M with that start and nothing on the end.
The Japan video has the same start, plus the 5 words on the end.
So your topic doesn't have to be exciting. Engineering isn't. What gets the click is the title: a boring subject, sold as something the viewer has missed.
For your own channel, swap in a different country and project:
"How [Country] Built the Impossible [Size or cost] [Project] that You've Never Heard Of"
Here's the prompt I'd paste into Claude (or ChatGPT) to find your first 20 video ideas:
"List 20 real engineering projects in [country or region] which look impossible and which most people have never heard of: huge scale, hostile terrain or a deadline nobody believed. For each, give the single hardest problem the engineers faced, 3 numbers with a source for each, and a title in this format: How [Country] Built the Impossible [Size or cost] [Project] that You've Never Heard Of."
Waiting for an exciting niche is wasted time.
Make the title the interesting part.
If you want me to tell you which project I'd start your channel with, DM me "BRIDGE"
Elevate retweeted
If your channel is dead, the content is probably fine.
The account was COOKED before the first real upload went live.
And you end up spending the next 3 months rewriting scripts, testing thumbnails and blaming the niche, while the real cause sits untouched….
So here's the checklist I run to find the real cause, in order:
1. Trust score. Nobody tested whether the channel gets views at all, so 15 finished videos went into an account the algorithm wrote off on day one. Fix: upload 1 boring 3 minute video and wait 72 hours. If it gets fewer than 300 views, abandon the channel.
2. Proxy fraud score. If your proxy's IP has a high fraud score, your reach dies, and YouTube never sends a flag or an email to tell you why. Fix: check the IP on browserleaks from INSIDE the browser profile before you upload anything. It takes 30 seconds.
3. Fingerprint coherence. Your browser profile tells YouTube what computer you're on, down to the processor and the graphics card. If those parts don't add up to a real computer, the account looks fake.
This one cost me a channel that had already passed the views test. Strong scripts, solid thumbnails, the right niche, and zero views for 3 weeks.
The problem was the profile. It said the computer had an Apple Silicon processor with an Nvidia graphics card, and no real computer has that combination.
I fixed the fingerprint, uploaded the same content, and it came back with over 800 views.
4. Upload cadence. 5 videos in 3 days on a fresh channel looks exactly like a spam farm. Fix: 1 upload every 48 to 72 hours for the first 2 weeks, then ramp up.
5. Niche contamination. Romance, then WW2, then survival. By video 10 the algorithm has conflicting data on 3 audiences, so it shows your videos to none of them.
See how content isn't even on the list? It's almost never the core issue.
One more thing: people constantly confuse a broken channel with a slow one.
Kill a channel fast if it fails the checks above and you can't fix it.
But if it passes them and is only growing slowly, leave it alone. YouTube needs roughly 12 to 15 videos to work out who your audience is, so quitting a healthy channel at video 8 can leave you up to 7 uploads short.
So work out whether your channel is broken or just slow before you change anything.
The full diagnostic and proxy setup is inside elevate(.uno)
Or DM me "REVIVE" if your channel is stuck at zero and you don't know why
Elevate retweeted
I have killed more channels than I have grown, dozens of them, and every dead one taught me the same lesson:
The algorithm evaluates a CHANNEL before it ever evaluates the videos it publishes.
Here is the 7 step framework built from those failures.
1. The impression test. Boring 3 minute horizontal video, film the floor, title nobody would search.
72 hours. 300+ impressions from browse or suggested means functional, 500+ means strong, 1,000+ means blessed.
If under 250, test another channel.
2. The fraud score. Your proxy IP carries ONE.
High score tanks channel trust with no warning. Check browserleaks from inside the anti-detect profile before uploading.
One IP per channel, never shared.
3. Fingerprint integrity. Your anti-detect browser reports what computer you are on. So if your profile claims a Mac but reports an Nvidia graphics card, there’s a chance you get in trouble with YouTube.
I lost a validated channel to this.
4. Posting cadence. Weeks 1 and 2, one upload every 48 to 72 hours, never batched.
Weeks 3 to 6, move to every 24 to 48 hours if the data looks healthy.
Week 7 onward, optimize on CTR and APV.
5. Upload timing. For English content hitting the 55+ US audience: 7 to 10am, 6 to 10pm, and 10pm to 1am EST.
Publishing at 2pm on a Tuesday means your test impressions land in a dead window and weak early data stops the algorithm scaling the video.
6. Playlists into podcasts. Group content into themed playlists, then convert the playlist into a podcast.
Session time climbs and you surface in YouTube Music feeds most faceless operators ignore completely.
7. The collaboration transfer. Invite one established channel in the same niche to collaborate on a new channel's video. Algorithmic trust transfers.
Once per new channel, no more, or you damage the strong channel's credibility in the eyes of YouTube.
Full walkthroughs for all 7 steps are inside elevate(.uno)
Or DM me "ALGO" if your channel is stuck at 0 views
Our community has generated $2M+ from AI story channels.
The same 6 myths still kill new channels every single month…
"I need a perfect thumbnail."
A thumbnail only matters after YouTube shows the video. A 3 word thumbnail you tested beats a 6 hour one you never shipped.
"Longer videos do better."
Longer videos don't earn more. Retention does. A tight 6 minutes beats a padded 12 almost every time.
"I should hide that it's AI."
Realistic AI generated or altered content needs YouTube's disclosure label where required. Hiding it is NOT a strategy. It's a policy risk. Read the current rules
"Copying a top channel is safe."
Copy the logic: structure, pacing, packaging. Never copy the script or footage. Reused content is a fast route to a monetization rejection.
"Good content always gets found."
Not if the title and thumbnail are weak. Nobody watches the video you're proud of if they never click it.
"If video one flops, the niche is wrong."
One flop is a data point. Retest with a different angle before you bury it. In our testing process, we wait 72 hours before reading anything.
Notice what's missing from the list.
Talent. Luck. A secret tool.
Most failures are bad beliefs, not bad creators.
Fix the belief, then fix the video.
Waiting to feel creative is the most expensive habit on YouTube.
"I'm not inspired today" is a hobby sentence.
Operators who last aren't more creative than you. They're more repeatable.
They have a topic source they check every week. A script template. The same voice. One upload slot on the calendar.
So video 14 doesn't need a spark. It needs the checklist.
Creativity picks the angle.
The system ships the video.
If your output depends on your mood, you don't have a channel. You have a mood.
Treat it like a business: fixed inputs, fixed process, data on every output.
Boring is the point. Boring is how you get to video 50.
Alt hooks:
If your uploads depend on your mood, you don't have a channel. You have a mood.
Nobody hit video 50 on inspiration.