How To Grow A Faceless Youtube Channel In 2026 | (ADVANCED COURSE)

[00:00] You're about to watch one of the most

[00:01] advanced YouTube automation courses on

[00:03] the internet. My name is Romero and I've

[00:05] been running Faceless YouTube channels

[00:06] since 2018. And since then, I've gotten

[00:08] hundreds of millions of views and I've

[00:10] made over $2 million in just YouTube ad

[00:13] revenue. In the next hour, you're going

[00:14] to see a handful of videos from my

[00:16] private school community. And the whole

[00:18] point of this is to see how many

[00:19] subscribers I get for giving out a lot

[00:22] of information as opposed to giving out

[00:23] some information and, you know, kind of

[00:25] plugging in my school community. Uh, I

[00:27] want to do the opposite now. I want to

[00:28] hit 200,000 subscribers. So, don't

[00:30] worry. There's literally no gatekeeping

[00:32] in this video. Just sit back, take

[00:33] notes, and hopefully you find

[00:35] information that makes you subscribe.

[00:36] All right, let's get started.

[00:40] So, the YouTube algorithm just had its

[00:42] biggest change ever. Well, kind of

[00:44] because it wasn't done on YouTube. Like,

[00:46] this isn't a an update that YouTube is

[00:48] going to go out and, you know, tell you

[00:49] guys that changed. This is more on the

[00:50] back end of Gemini, and it was more of a

[00:53] developer update as opposed to like a

[00:55] creator update. Uh, so that's why you're

[00:57] probably not going to hear anything

[00:58] about it. Now, the way the old algorithm

[01:00] worked, you posted a video with high

[01:02] click-through rate, which was clickbait,

[01:03] or, you know, just a good thumbnail and

[01:05] title. And then it would get clicks.

[01:07] From there, the algorithm would decide

[01:09] based on average view duration if it was

[01:11] going to continue pushing it, which if

[01:12] you had high CTR, high AVD, you would

[01:15] obviously get more of a push. And then

[01:16] this cycle would repeat until the end of

[01:18] the video cycle. And of course, tags and

[01:21] descriptions and keyword stuffing would

[01:23] make this the meta and just, you know,

[01:24] make it 10 times more powerful. But with

[01:26] this came the spam channels, the AI

[01:28] channels, the rewriting scripts, taking

[01:30] people's content and just rewarding it.

[01:32] There was a lot of duplicates. There was

[01:34] people exploiting this. And essentially,

[01:36] the system no longer worked because of

[01:37] fake spam and abuse. So, Google's idea

[01:40] was to uh involve Gemini. So, every time

[01:42] you post a video, Gemini reads the

[01:44] description, the transcript of your

[01:46] video, and the SEO. And from here, it

[01:48] decides if it gets flagged or if it gets

[01:49] approved. Now, in the beginning of

[01:51] January 2026, literally a few days ago,

[01:54] Google updated Gemini. And what this

[01:55] does is that now it gave Gemini the

[01:57] ability to reason. Meaning that it can

[02:00] see exactly what you search for on

[02:01] Google, on Google Drive, on Google Docs,

[02:04] Gmail, uh, and YouTube. So, this is what

[02:06] the meaning of of reasoning is. It can

[02:08] now see everything that you do across

[02:10] all of the Google platforms. Um, and

[02:12] it's kind of like Facebook. It's kind of

[02:14] more like targeted ads. So, a video is

[02:16] no longer based on just the keywords,

[02:18] the the metadata, and all that, right?

[02:20] Because now Gemini can reason if it's a

[02:23] video to recommend. And what it goes

[02:25] based off of is called semantic ID. This

[02:28] is the biggest change to Google's

[02:30] algorithm overall. What a semantic ID is

[02:33] is when you upload a video, Gemini now

[02:35] sees the editing speed of your video,

[02:37] the colors you use, the volume, the

[02:40] energy, and more specifically the

[02:41] intention behind the words that you are

[02:43] using in your script. So when Google

[02:45] analyzes this, it gives it a semantic

[02:47] ID, which is, you know, a high energy

[02:49] video. Let's say it's a gaming video,

[02:50] it's a male voice, it's a chaotic edit.

[02:53] And this is like the perfect

[02:54] description. It's a code for your vibe,

[02:57] the vibe of your video, the energy of

[02:59] it. So, this is one of the most advanced

[03:01] ways that the algorithm now recommends

[03:03] content. So, let's say a user is on

[03:05] Google and they're searching Minecraft

[03:06] or Fortnite. So, Gemini from here is

[03:08] assuming that the person is in the mood

[03:10] to watch content with high energy,

[03:11] right? Obviously, gaming focused cuz

[03:13] it's it's it's gaming videos, but you

[03:15] know that that that stereotypical

[03:17] YouTuber that's yelling into the mic and

[03:19] you know, very very excited. Typical

[03:20] gaming channel. So Google tells Gemini,

[03:23] "Hey, they're searching Minecraft and

[03:24] Fortnite. What's the next logical thing

[03:26] that this person wants to watch? What

[03:28] what are they in the mood for at this

[03:30] current moment?" So Gemini understands

[03:31] that the person wants to watch high

[03:33] energy content, gaming content. That is

[03:35] their mood. That is their vibe. And it

[03:37] starts thinking of two options. Does it

[03:39] recommend you a gaming video with high

[03:41] energy? Or does it recommend you a

[03:43] gaming video that is storytelling and

[03:45] slow pace? You know, the answer is

[03:46] pretty obvious. it's going to recommend

[03:48] the video with high energy and high pace

[03:50] because if you're watching Fortnite and

[03:52] Minecraft videos, typically it's very

[03:54] high energy YouTubers, but if the user

[03:56] was googling content about Red Dead

[03:58] Redemption, which if you don't know,

[03:59] it's a very storydriven uh video game.

[04:01] It would then choose the slowpaced story

[04:04] to recommend, right? Because it matches

[04:05] the tempo and the vibe that the person

[04:07] is feeling. It's no longer about just

[04:09] the topic. So, this is what makes the

[04:11] win emotion-wise. It's the fast rapid

[04:13] cuts, the high dopamine state. audio is

[04:16] loud and shouting matches intensity

[04:18] sentiment competitive and high stakes

[04:19] and you know it recommends the Fortnite

[04:21] video based on that it is the energy

[04:23] signature the true semantic ID now

[04:26] that's for entertainment videos like

[04:28] gaming right this is for people that are

[04:29] trying to make um you know very

[04:31] emotional videos but for us you know the

[04:33] information niche that is the celebrity

[04:35] news that is all of the niches that we

[04:37] do the algorithm sees it different it

[04:39] doesn't just go off of energy because

[04:41] that's not a fair meter so different

[04:43] goals equals different match matching

[04:45] which is no longer mood match it is now

[04:47] niche match. So Google has a knowledge

[04:49] graph right and every topic has its own

[04:52] web which is called nodes. For example

[04:54] Elon Musk the name Elon Musk on Google

[04:57] has this specific code here M03_D0.

[05:02] A node would be Tesla M07Z_3S.

[05:06] Another node would be Neurolink. Another

[05:09] one would be Mars. Another one would be

[05:10] SpaceX. Now what Google sees is not

[05:13] specifically SpaceX. They see this code

[05:15] and this code represents SpaceX across

[05:18] all of Google. So the idea now is to hit

[05:21] every single node on the topic of your

[05:23] video. So our video script for semantic

[05:26] seating would be in today's video about

[05:28] Elon Musk. There's one. We'll dive into

[05:30] the latest news with Tesla Cybertruck.

[05:33] There's two. The progress of SpaceX's

[05:35] Starships. That is three. The future of

[05:37] Neurolink brain chips. That's uh node

[05:40] number four. and the ultimate goal of

[05:41] colonizing Mars. That is the fifth node

[05:44] that is hit. So when you hit the nodes,

[05:46] you increase the authority of your video

[05:48] because you cover absolutely everything

[05:50] in Google's knowledge graph. 99.9% of

[05:52] YouTubers do not know this. The only

[05:54] people that know this is like Google

[05:56] engineers and back-end algorithm people,

[05:58] right? But when it comes to creators and

[05:59] when it comes to YouTubers, we're

[06:01] focused on like, oh, high AVD retention,

[06:04] how long people watch. Nobody is

[06:06] thinking of Google's knowledge graph and

[06:08] hitting every single node. This one here

[06:10] is the most important one, the

[06:12] originality one, because Google is now

[06:15] going based off of net information gain,

[06:18] which is uh the main purpose of of this

[06:21] update, I believe. So, when you post a

[06:22] video, um, and if you've been around for

[06:24] a while, you know that you're going to

[06:25] have probably like 20 clones, someone

[06:27] that rewrites your script and reposts

[06:29] it. This is uh the sea of sameness,

[06:32] commodity content. So, if you recreate

[06:35] someone's video, you're just going to be

[06:36] added to this huge sea of replicated

[06:39] content that Gemini says, "I've seen it

[06:42] before. I have this in my knowledge

[06:43] graph." What Google wants from you is

[06:45] for you to give it new information.

[06:47] Either that's personal, either that is

[06:49] focusing your video on something very

[06:52] specific, which is micro niching, or by

[06:54] approaching it with a contrarian point

[06:56] of view. So, your video's total info

[06:59] cannot be in the existing Google

[07:01] knowledge base. So if you say something

[07:03] original or if you approach something in

[07:05] in an original way, you add net

[07:08] information gain, something that Google

[07:10] has never seen before. So when you post

[07:12] a video, Gemini starts seeing, okay, I'm

[07:14] going to rank this video against the top

[07:16] 10 and I'm going to see if there is

[07:19] enough new information that I uh that I

[07:22] can see. So are you adding to the pile

[07:24] or are you changing it? Are you

[07:26] approaching it in a different way? And

[07:27] if you are approaching something in an

[07:29] original and different way, uh you get

[07:31] rewarded for that. Which is why the

[07:33] originality score is so important. And

[07:34] not only that, if you make a script and

[07:36] you get an originality score of a 10 out

[07:39] of 10 and you hit every single semantic

[07:42] ID of that topic, Google is going to see

[07:45] your video as high authority and it's

[07:47] going to understand absolutely

[07:49] everything about your script. By the

[07:50] way, this whole time we were [ __ ] up

[07:52] with our scripts. And the perfect

[07:53] example was that uh Gen Z consumer

[07:56] behavior. Logically, we would think,

[07:57] okay, by saying young woman, you would

[08:00] assume that Google understood that you

[08:02] were talking about Gen Z. But no, you

[08:04] have to verbally say it. So the goal is

[08:06] now to create a what we call delta,

[08:09] which is essentially your unique value.

[08:11] For example, if you were to make a video

[08:13] about weight loss, Google's knowledge

[08:15] graph already has exercise, diet,

[08:17] calories, metabolism, right? every every

[08:19] single video about weight loss has those

[08:21] four key points. But if you want to hit

[08:23] that net information game, you would

[08:25] talk about brown fat activation and then

[08:28] with the semantic ID code here at the

[08:30] bottom, that would be new data for

[08:32] Google, a new approach. So you can do

[08:34] this with information, right? And if you

[08:36] want to know if Google has something in

[08:38] their knowledge graph, it's going to

[08:39] tell you right here in the originality

[08:41] score. And and also you could go with

[08:42] the easy one, which is the contrarian.

[08:44] You know, say the opposite of what

[08:46] everyone says. If someone says, "Oh, I

[08:48] like Starbucks coffee because it's

[08:49] better than, you know, [ __ ] Dunkin

[08:51] Donuts." You're going to say, "Actually,

[08:52] no. Dunkin Donuts is better." That's a

[08:55] hypothetical, right? That would give you

[08:56] the contrarian aspect, which also in

[08:58] itself is a new opinion that Google can

[09:01] can essentially store. And also very

[09:03] important for you to note, we are the

[09:05] first ones that know about this. Nobody

[09:07] else, no other YouTube community, no

[09:09] other Google community is going to tell

[09:10] you about this. And I'm probably not

[09:12] going to post it on YouTube yet cuz I

[09:14] know the comments are going to be filled

[09:15] with a bunch of bunch of dumb asses that

[09:17] are going to be like um um that's not

[09:19] true. YouTube didn't announce it. All

[09:21] you need to do is just make highquality

[09:23] content. You know that [ __ ] It's

[09:25] like no. You know, now this doesn't

[09:26] guarantee success obviously, but it is

[09:29] coding your script to meet every

[09:32] requirement that Google is looking for.

[09:34] Now, another way to to build net

[09:36] information gain or to give net

[09:38] information gain to Google is to connect

[09:40] two things unrelated. Example,

[09:43] connecting AI to agriculture. So, when

[09:45] the algorithm tries to predict what the

[09:47] transcript of your video is, because

[09:49] it's always done that, um, that's kind

[09:50] of how you get caught with reused

[09:51] content, right? If the algorithm or

[09:53] Gemini could predict the next word of

[09:55] your script consistently, it gets hit

[09:57] with reused content because it's like,

[09:58] okay, this was clearly made with AI. But

[10:00] if you break the algorithm's prediction

[10:02] path with high net gain, um you do a

[10:05] pattern interrupt and Gemini is like,

[10:07] "Bro, this is fire, right? I have never

[10:09] seen this before." So right now in the

[10:11] sea of sameness, everyone that is trying

[10:14] to recreate the same style of videos,

[10:15] the same content, rewrite scripts, if

[10:17] you approach every single script with a

[10:19] strategy, the reward should be priority

[10:21] recommendation. You know, with this new

[10:23] Google update, we literally cracked it.

[10:24] You know, everyone's going to freak out

[10:26] about this if I say it publicly. um

[10:28] they're going to think like, "Oh, the AI

[10:29] is watching everything. It's over." It's

[10:30] like, "Not really. Now it's kind of

[10:32] easier because now we understand exactly

[10:34] what it's looking at." And it's it's it

[10:35] was kind of in front of us all the time.

[10:37] Now the game is to be completely

[10:39] unpredictable. Show the algorithm

[10:40] something it's never seen before cuz

[10:42] Google knows everything. I mean, you

[10:44] know, imagine how many repeats it sees a

[10:46] day. And it goes back to the main point

[10:47] that Google is trying to become an

[10:49] answer engine instead of a search

[10:52] engine, right? And if you Google

[10:53] anything, for example, if I Google

[10:55] myself, there's an AI mode for

[10:57] everything. And sometimes when you

[10:58] Google something alone, you see the AI

[11:00] overview before you see, you know,

[11:02] actual search terms because Google wants

[11:04] to have the answer for everything. They

[11:05] no longer want people to be scrolling

[11:07] through and and looking for for their

[11:09] answer. They just want to give it to

[11:11] them straight up. So, by adding new

[11:12] value, giving it new information that it

[11:14] could potentially use for AI overviews,

[11:16] which is the ultimate goal of Google,

[11:18] that's how we're going to win in in

[11:19] 2026. So yeah, there's a few other

[11:21] things I'm going to mention within this

[11:22] update. You know, when when Google uh

[11:24] made this update with Gemini, it's it

[11:26] sent a ripple effect throughout YouTube

[11:28] and how it works. So yeah, keep an eye

[11:29] out for those videos. It's going to it's

[11:31] going to be coming up pretty soon.

[11:35] So this is going to be one of the most

[11:37] important videos in the entire course.

[11:39] If for some reason this YouTube

[11:41] automation journey or YouTube journey

[11:42] fails for you, you can still take this

[11:44] with you. And that's because what I'm

[11:46] about to teach you right now applies to

[11:47] absolutely everything with social media.

[11:50] Doesn't matter if it's marketing,

[11:51] doesn't matter if it's building your own

[11:52] personal brand, doesn't matter if it's

[11:54] faceless YouTube channels. This is

[11:55] across the board the most important part

[11:58] with Google and YouTube in 2026. And no

[12:01] guru or no coach is ever going to tell

[12:04] you this. And again, that's because they

[12:05] don't know. As someone who has been

[12:07] doing YouTube for years, as someone who

[12:09] runs multiple channels, as someone who

[12:11] has spoken to thousands of people, I

[12:14] know when the algorithm changes, I know

[12:16] when something happens. I can tell the

[12:18] smallest detail. It's part of my like

[12:19] Asberers, okay? I I I get really um I'm

[12:23] really good at noticing slight changes

[12:26] with the algorithm. And I'm not saying

[12:27] that like in an ego way, like I'm a like

[12:29] I'm a [ __ ] genius, right? I'm

[12:30] terrible at 99% of the things in life,

[12:32] but noticing algorithm changes is

[12:34] probably the only thing I'm good at. And

[12:36] then towards the end of 2025, I realized

[12:39] that a change had happened, but it

[12:42] wasn't like the others. This was

[12:43] actually a very big change, one of the

[12:45] biggest ones. A few months later, Google

[12:47] documentation came out. And in fact, it

[12:49] was it was one of the biggest changes

[12:51] that Google and YouTube have done. Now,

[12:53] if you're starting a new channel or

[12:54] you're continuing a channel, it doesn't

[12:56] matter where you are in your journey,

[12:58] okay? This is something that you have to

[13:00] do. And uh I guess the the whole point

[13:02] of this video is offplatform presence.

[13:05] Okay, you're invisible to Google. And

[13:07] for those of you that don't know, Google

[13:09] owns YouTube. They are are now as of

[13:11] today, March of 2026, sharing almost an

[13:14] identical algorithm and it's all

[13:16] connected. Okay, Google bought YouTube.

[13:18] So now they're both kind of the same

[13:20] [ __ ] So you're invisible to Google.

[13:21] Where are you in Google's knowledge

[13:23] graph? In 2024, there was a Google API

[13:27] leak where thousands of internal Google

[13:29] documents were leaked. And something

[13:30] that was in those Google documents was

[13:33] something called Nav Boost, which

[13:35] essentially admitted that Google would

[13:38] push branded channels over unbranded

[13:41] YouTube channels. Now, by branded, I

[13:43] don't mean a cool logo and, you know,

[13:45] cool watermarks in your video. That's

[13:46] not what I mean. The first thing you

[13:48] want to do right now is create a

[13:50] presence outside of YouTube. And here is

[13:53] the key difference. an entity versus an

[13:54] uploader, right? Someone that just

[13:56] uploads on YouTube has low trust because

[13:58] they don't appear anywhere else. But an

[14:00] entity would be someone that is on

[14:03] YouTube, they're also on Indeed, they're

[14:05] also on Pinterest, they're also on

[14:06] Facebook, they're also on Instagram,

[14:08] they're also on X or Twitter. And all of

[14:10] them connect back to the YouTube

[14:11] channel. They have the same name and

[14:13] they have the same branding. Now, by

[14:14] having all of these things connected,

[14:16] you don't just become a YouTuber now,

[14:17] right? You become an entity, which

[14:19] increases your trust score times 100.

[14:22] Okay, this is one of the most important

[14:23] parts with the algorithm today because

[14:25] in 2026, anybody can start YouTube

[14:28] channels with AI. It has AI has made it

[14:31] extremely easy to launch channels and AI

[14:34] has also made it extremely easy to start

[14:36] content farms, which is just, you know,

[14:38] having AIs and n automations posting 50

[14:42] videos a day. And YouTube knows that.

[14:44] And to combat it, what they're doing now

[14:45] is that they don't necessarily care what

[14:48] you're uploading. They care about who is

[14:50] uploading it. So for that you need to

[14:52] become an entity. And the way this works

[14:54] is someone sees our Pinterest, they go

[14:57] to Google, they search us up and our

[14:59] authority goes up. And when this is

[15:01] repeated in every single platform, the

[15:03] authority grows more. So if I was to go

[15:05] and search up Views for Income LLC on

[15:09] Google, you're going to see that I have

[15:10] a Google knowledge graph here. Right? So

[15:12] this means that Google sees Views for

[15:15] Income LLC as a legit company. I have

[15:18] the phone number which I don't know

[15:19] whose phone number this is, the address,

[15:21] you know, location, whatever, the

[15:22] website and the description here. Now,

[15:25] this is ultimately what you want for

[15:27] your YouTube channel. So, the action

[15:29] plan is to add same ass schema to your

[15:32] site. This is going to be a completely

[15:33] different video. This is more of an

[15:35] introduction. You want to link all of

[15:36] your profiles with each other. You want

[15:38] to drive traffic from outside of YouTube

[15:40] and become a net contributor. This is

[15:42] the sauce, right? Especially with all of

[15:44] the demonetization going on nowadays,

[15:46] YouTube won't demonetize you if there is

[15:49] enough proof that you are a entity, that

[15:52] you are a brand and not just a random

[15:54] YouTuber. Now, there is a lot more that

[15:56] you need to know about this, which I'm

[15:57] going to cover in the next video. Uh

[15:59] before we do that, you want to do me a

[16:00] favor, guys, search up Romero, go to

[16:02] this Reddit tab here, cuz Reddit

[16:04] apparently has a lot of weight. You

[16:05] know, go in here, read some of the

[16:07] comments, leave a positive comment about

[16:09] me if you can. Would appreciate that, my

[16:11] boy. But yeah, once you start a YouTube

[16:12] channel, don't forget to make an

[16:14] Instagram, Facebook, and Twitter with

[16:16] the same name as your YouTube channel.

[16:18] Make sure that you link your YouTube

[16:20] channel in the bio of each. And make

[16:22] sure that in your YouTube channel, you

[16:23] also link all of those things. And I'm

[16:25] going to show you how to do that real

[16:26] quick. So, go to YouTube, click on the

[16:28] top right profile picture, go to view

[16:31] your channel, then go to customize

[16:33] channel here at the top, scroll down,

[16:35] and you're going to see links. Share

[16:38] external links with your viewers.

[16:39] they're visible on your channel's

[16:41] profile and about page. So then you

[16:42] would click add link from here. You

[16:44] would do Instagram, put your URL there,

[16:47] add another one, do your Twitter. So

[16:49] yeah, go ahead and start doing that. I

[16:51] think this is a pretty good pretty good

[16:52] intro for now on on what you're about to

[16:54] see next.

[16:58] All right, so I'm going to break this

[16:59] down again. And this is like uh the how

[17:01] to problem solve your channel. And I

[17:02] know this was covered in another video.

[17:04] I already broke it down. I already

[17:05] showed you guys how it is. But I feel

[17:06] like there was a some sort of

[17:08] misunderstanding. Okay. So, what I'm

[17:10] going to do is pull up some screenshots

[17:12] of what students have sent me for the

[17:14] channel reviews during the the live

[17:15] calls and I'm going to show you like so

[17:17] you can see my perspective of the of the

[17:19] [ __ ] that I see. Right. Here's

[17:20] screenshot number one. Now, this guy in

[17:22] the last 28 days has gotten 9.9,000

[17:24] views, 166,000 impressions, 5.2% CTR,

[17:28] and a 3 minute and 50 second AVD. Right?

[17:31] And he's been sort of consistent here.

[17:33] Uh whatever he posted on May 9th, that

[17:34] popped off. probably a diddy to be

[17:36] honest. And then after that it kind of

[17:37] slowed down, picked up a little bit. Um,

[17:39] and the peak that he got was 1,900 views

[17:42] in a single day. And now 166,000

[17:45] impressions is not a crazy amount for

[17:48] the last 28 days, but what is bad is the

[17:51] views he got. And you see this and and

[17:52] what the what I see here is like, okay,

[17:55] YouTube completely ignored his channel

[17:57] uh at the end of April, beginning of

[17:58] May. And then on May 9th, it took it

[18:00] into consideration and it tested his

[18:02] channel, right? So then every single

[18:03] time he was posting, uh, obviously he

[18:05] was getting some peaks. YouTube was

[18:06] testing it out and this student was

[18:08] asking me, "I don't understand what I'm

[18:09] doing wrong." Like, "Bro, your CTR is

[18:11] 5.2

[18:13] and your AVD is 350. What the [ __ ] do

[18:15] you mean you don't know what you're

[18:16] doing wrong? It's so simple. It it it's

[18:19] so it's right there, right? It's right

[18:20] there for you. It's a 5.2% CTR. And yes,

[18:23] this guy's in celebrity news. He was

[18:24] posting Diddy videos. You have a 5.2%

[18:27] CTR." And I asked him during the call. I

[18:29] was like, "Well, okay, your thumbnails,

[18:30] you know, they're not the worst. They're

[18:32] different, but they're not the worst.

[18:33] You can switch to to the normal style of

[18:35] thumbnails that I teach in in the

[18:36] course. And also, how do you do your

[18:38] titles? And he said, "Oh, I use Chad

[18:39] GBT." Bro,

[18:45] you know, guys, for celebrity news, you

[18:47] can't use Chad GBT for titles. You

[18:49] can't. It's not going to be as good. And

[18:50] that that's not how this works. Yes, you

[18:52] can use Chad GBT for titles with

[18:54] literally any other niche. You know,

[18:55] anything that's evergreen, cool. But

[18:57] celebrity news, it can't be that way.

[18:58] Remember, we have to beat our

[19:00] competitors with creativity, right? So,

[19:01] a 5.2% CTR. YouTube tested it. He didn't

[19:05] impress YouTube with the CTR, but not

[19:07] only that, it's a 3 minute and 50 second

[19:09] ABD, which is not the worst that I've

[19:13] seen. Okay, it's not the worst, but it's

[19:15] also not great. So, not only did he not

[19:17] have the bare minimum of CTR that you're

[19:19] supposed to have, I mean, a 9% is like,

[19:21] okay, if you had a 9% CTR and you did

[19:24] and these were your analytics, I promise

[19:26] you, you would have way more views,

[19:27] right? If you're going to have low CTR,

[19:29] you better have [ __ ] high ABD. And

[19:31] same goes for ABD. If you have low AVD,

[19:35] you better have really high CTR. Okay?

[19:37] Vice versa. Same [ __ ] And this guy

[19:39] didn't have any of them. So, it's like,

[19:40] okay, well, here's time to problem

[19:42] solve. Number one, stop using ChatGBT

[19:44] for your titles, right? Use a little bit

[19:46] of creativity. You know, again, put

[19:48] yourself in the perspective of the

[19:50] audience. Um, and that should fix. Now,

[19:51] for Evergreen, that's a completely

[19:53] different story. I'm going to cover that

[19:54] in other videos, but for celebrity news,

[19:56] this is unacceptable. Okay, for AVD,

[19:58] same thing. And sometimes it's the

[19:59] smallest changes that you can do. It can

[20:01] be the background music. It could be the

[20:03] voice you're using because a lot of

[20:04] students have shown me their channels. I

[20:05] I sit there and I watch. I'm sure if

[20:07] you've joined the live calls, you can

[20:08] see that the voices are dog [ __ ] I can

[20:10] tell you're not using 11 Labs. Some of

[20:12] them are okay. Some of the voices that

[20:14] that are being used are okay, but the

[20:16] majority of them sound so bad. And I

[20:18] always have to say, and if you sit

[20:19] during those calls, I'm like, "Okay, cut

[20:21] the gaps in between the audio. Okay,

[20:23] change voice." Like, guys, yes,

[20:24] celebrity news can be dog [ __ ] like

[20:27] quality of the content, but come on now,

[20:29] you know, there, come on now. It can't

[20:31] be that bad. And this guy, you know, he

[20:33] wasn't shadowbanned. It took a while for

[20:35] YouTube to to push his content, but when

[20:38] YouTube gave him the opportunity, he he

[20:40] fumbled. That's what he did. Next one is

[20:43] this. I believe it's the same guy. Last

[20:45] 7 days, 16,000 impressions, 972 views,

[20:49] 5.1% CTR, and an average view duration

[20:51] of 2 minutes and 33 seconds. At this

[20:54] point, when you see these numbers, guys,

[20:56] you have to problem solve something,

[20:58] right? You can't just keep posting and

[20:59] posting and posting and hopefully, you

[21:00] know, it's going to pop off, which, yes,

[21:02] it can happen. It's very likely that one

[21:04] video is going to pop off regardless of

[21:06] what these numbers look like. But

[21:07] problem solve, you know, if your AVD is

[21:09] 2 minutes and 33, add some phase 2 twos

[21:11] at the beginning. Maybe remove some

[21:13] phase twos if you have them. Change the

[21:15] voice, change the background music, cut

[21:16] the gaps of the voice, switch up the

[21:18] script a little bit, right? Change

[21:20] change your prompt. do something because

[21:22] 2 minutes and 33 seconds of ABD is

[21:24] unacceptable. That's like a [ __ ]

[21:26] YouTube short. So, you have to problem

[21:28] solve. It's always going to be the

[21:29] dumbest, smallest thing. Just change

[21:30] stuff. Change stuff every single video.

[21:32] Instead of just repeating and sending

[21:33] the same [ __ ] out, change something of

[21:35] it, post it, and then analyze, right?

[21:38] CTR. I mean, if if you just don't got

[21:40] it, you don't got it. It's one of those

[21:41] things, and I mentioned that in in the

[21:42] course as well, the the VIP. If you

[21:45] don't have the creativity where you can

[21:46] sit down, put yourself in the in the

[21:48] shoes of the audience and come up with

[21:49] something creative, something original,

[21:50] continuing the story, um, for these

[21:52] celebrity news niche, you know, some

[21:55] people just aren't built for it, man.

[21:56] Some people I just can't teach. You

[21:58] know, it's not something that you you

[22:00] teach. I showed you guys how I did it in

[22:01] one of the VIP videos. I showed you how

[22:03] I came up with the titles, putting

[22:04] myself in the shoes of of the audience,

[22:06] and, you know, for some people that just

[22:08] doesn't click. That doesn't mean you

[22:09] can't, you know, your YouTube automation

[22:10] career is over. That just means the

[22:12] niche isn't for you. Or at least that

[22:15] subniche, you know what I mean? Like,

[22:16] okay, maybe Diddy won't hit for you. You

[22:18] know nothing about Diddy. You knew it

[22:19] from the jump. It's not going to hit for

[22:21] you. Okay, well, go for old Hollywood.

[22:23] You know, go for any of the I put so

[22:25] many evergreen niches in there and

[22:26] subniches of celebrity. And you have a

[22:28] lot of options, right? For those, you

[22:30] don't need to be that creative. Next one

[22:31] is this guy. Last 28 days, 23,000 views,

[22:35] 362,000 impressions, almost double from

[22:37] the other guy. 5.2% 2% CTR 3 minute 49

[22:41] second AVD and then he stopped posting

[22:43] like bro peak views he got looks like uh

[22:46] 4.3 4.5 right here didn't adjust

[22:50] anything YouTube was begging this

[22:52] [ __ ] to fix something and to

[22:54] adapt his CTR adapt his AVD change

[22:57] something and he didn't and then he just

[23:00] stopped posting and the second I'm

[23:02] assuming he posted something here but

[23:03] deleted it picked up you know when he

[23:05] posts it picks up but when he doesn't it

[23:07] dies Look. So, every time he posts it,

[23:10] it peaks. So, this is a good channel.

[23:11] This is a really, really good channel.

[23:12] He's just not fixing things. It's about

[23:14] problem solving, guys. This is YouTube

[23:16] automation is just pattern recognition.

[23:18] It's just pattern recognition. It's all

[23:20] it is. Okay? So, if if you're going to

[23:22] send us a channel review and you're

[23:23] like, I don't understand what I'm doing

[23:24] wrong. I don't understand why the

[23:25] channel isn't working. And you show me

[23:27] these [ __ ] numbers, dude, I'm going

[23:28] to be like, come on. It's the AVD or

[23:31] it's the CTR, right? Change something.

[23:33] Change the approach. Right? And you can

[23:35] I'm not saying don't send channel

[23:37] reviews for that reason. And if if you

[23:38] need more ideas on what you can change

[23:40] or what exactly would I start with, for

[23:42] sure, send them. But the answer is very

[23:44] obvious, right? This isn't rocket

[23:45] science. It's very simple. If your if

[23:47] your channel looks like this and you

[23:48] have the these amounts of uh

[23:50] impressions, YouTube wants this channel

[23:52] to work. You're just not letting it. The

[23:54] only issue that you should have is like,

[23:55] okay, your CTR is like a 9 plus and and

[23:58] your AVD is like 4 minutes and you just

[24:00] want to know how to how to reach five

[24:02] minutes in AVD or 15 minutes in CTR,

[24:05] right? those little micro adjustments.

[24:07] It's this is what you're going to

[24:09] benefit with the most. And yes, I'm

[24:10] going to be covering that a little bit

[24:12] more in in these videos. Cool. Just

[24:14] wanted to let you know if we're all on

[24:15] the same page and what I'm seeing with

[24:17] the majority of the students and what

[24:19] you're probably seeing and then we're

[24:20] going to problem solve this together in

[24:21] these videos before we go into the

[24:23] channel reviews. Okay, awesome. Let's

[24:25] move on.

[24:28] So, if you hear anyone say that average

[24:31] view duration is the most important part

[24:32] of the algorithm and that your ultimate

[24:35] goal should be to have someone watch

[24:36] your video for as long as possible, that

[24:39] is a lie. Sort of kind of because with

[24:42] the new Gemini update, that did change

[24:45] the way that YouTube perceives average

[24:47] view duration. So before, if you had a

[24:50] 10-minute video and you had a 2-minute

[24:52] average view duration, immediately the

[24:54] algorithm would be like, "This video is

[24:56] dog [ __ ] People left after 2 minutes.

[24:58] They didn't watch the full 10 minutes.

[24:59] Therefore, I'm not going to recommend

[25:01] this video." Which makes sense. That in

[25:03] itself is logical. But now that Gemini

[25:05] has the ability to reason and to

[25:07] understand what the viewer is doing

[25:10] before and after it clicks the video. So

[25:12] then Gemini goes and investigates, why

[25:14] did the person leave at 2 minutes and

[25:15] didn't watch the entire 10-minute video?

[25:17] Because again, Gemini now watches what

[25:19] you do before and after you view a

[25:21] video. So it asks, did the viewer leave

[25:23] because you got the answer? That would

[25:25] be good abandonment. The viewer is

[25:27] satisfied. Or did the person leave

[25:29] because they were bored or clickbait?

[25:32] That would be bad abandonment. And it

[25:34] knows this because let's say your video

[25:36] title is how to change a tire. Bad

[25:38] abandonment would be if the viewer

[25:40] clicks off of your video and starts

[25:42] searching for another YouTube video that

[25:45] is a tutorial on how to change a tire.

[25:47] This tells Gemini that the problem of

[25:49] the viewer was unsolved. Now, good

[25:51] abandonment would be when the viewer

[25:53] watches after 2 minutes and then either

[25:55] closes the app or starts searching for

[25:57] other content on YouTube saying that

[25:59] your tutorial video on how to change a

[26:02] tire. You know, you you teach people how

[26:04] to do it within 2 minutes of the video

[26:06] and then the rest is just like a vlog.

[26:08] Completely irrelevant, right? People

[26:09] don't care about that. They were just

[26:10] there to learn how to change a tire. So,

[26:12] that would be good abandonment if they

[26:14] leave and they don't search for that

[26:15] topic again. So again now that Gemini

[26:18] has not a conscious but you know it can

[26:20] reason it can now fully go forth with

[26:23] Google's objective which is uh

[26:24] satisfaction. Did it satisfy the viewer?

[26:26] Did it answer their question? And the

[26:28] more you get this right with Gemini

[26:30] understanding this now um that builds

[26:32] channel authority for you. Meaning that

[26:34] even if you get a 2-minute average view

[26:35] duration for 20 videos back to back if

[26:38] the viewer leaves happy you're going to

[26:40] be recommended every single time. That

[26:42] is impressions and that is views. So

[26:43] yeah, I mean average view duration is

[26:45] still important, right? There there are

[26:47] some niches um that you guys are in

[26:49] which are more like storybased I suppose

[26:52] where you kind of have to have a high

[26:53] AVD because anything else besides that

[26:56] would be the person got bored and left.

[26:58] But it's still too early to tell. Um at

[27:00] least we know that Gemini is now going

[27:01] to start doing this. So um not every

[27:03] metric is bad. That's why some of you

[27:05] see like a two 2 minute 30 average view

[27:07] duration. Uh but you get a [ __ ] ton of

[27:09] views. So yeah, again, if we think about

[27:11] this long term and we think about where

[27:13] YouTube is headed, not bad. Not bad at

[27:16] all, right? But yeah, let's see what

[27:18] Gemini does and hopefully this starts

[27:20] building a little bit more of an answer

[27:21] for us, a little bit more of an accurate

[27:23] answer. Um, so yeah,

[27:27] now before you start a faceless YouTube

[27:29] channel, there is a very very important

[27:31] question that you have to answer and

[27:33] that is, are you going into a trendy

[27:35] niche or are you going into an evergreen

[27:38] niche? both have their pros and cons.

[27:40] Now, if you already have an idea of what

[27:42] you're going to do, I want you to pause

[27:44] for a second um and hear me out, okay?

[27:46] Because you might not like what you're

[27:47] about to hear. A trendy niche would be

[27:50] anything that revolves around news. You

[27:53] know, for example, celebrity news. An

[27:55] evergreen niche would be documentaries,

[27:58] top 10s, you know, things that aren't

[28:00] time-sensitive. Because if you go into a

[28:02] trendy niche like celebrity news, you

[28:04] have to consider that you're going to

[28:06] have to be on top of this [ __ ] right?

[28:07] If you're going to be doing uh news

[28:09] about rappers, you have to be checking

[28:11] TMZ every single day. You have to be

[28:13] checking what the competitors are doing

[28:15] every single day. But most importantly,

[28:17] you're going to have to be uploading

[28:19] almost every single day. As opposed to

[28:21] an evergreen YouTube channel where you

[28:23] do history documentaries about World War

[28:25] II, you can get away with uploading two

[28:27] to three times a week. And you're not

[28:28] necessarily pressured about getting a

[28:30] video out. For example, let's go ahead

[28:31] and say that uh Jay-Z and Beyonce get

[28:34] divorced. You're going to have to post

[28:36] as soon as possible. Well, you're going

[28:37] to have to drop everything you're doing,

[28:38] right? And post about the divorce. But

[28:40] in the evergreen niche, you know,

[28:42] there's no pressure. You don't have to

[28:44] be first to anything. There isn't uh,

[28:46] you know, trends. Now, let's talk about

[28:48] money. With a trendy niche like

[28:50] celebrity news, again, using the same

[28:52] example of Beyonce and Jay-Z getting a

[28:55] divorce, that is gold, right? If you

[28:57] start posting videos, pumping out two to

[28:58] three videos that same exact day, and

[29:00] the next day you do it again, that month

[29:02] you're probably going to make a [ __ ] ton

[29:03] of money. The issue is that when nobody

[29:06] cares about Beyonce and Jay-Z's divorce,

[29:09] the views go down a lot. So, you're

[29:11] going to have to be on your toes 247.

[29:12] But in the evergreen niche, it's slow

[29:14] and steady. You post a documentary video

[29:16] and it doesn't go crazy the first day or

[29:18] the first week, but it slowly starts

[29:20] getting views. Let's say 200 views a

[29:22] day. Then you post another video, same

[29:24] thing. 200 views a day, starts growing

[29:26] slowly until eventually one of them

[29:28] picks up to like, let's say, 10,000

[29:30] views a day. and another one picks up to

[29:32] 10,000 views a day. Now you have two

[29:34] videos getting 10,000 views a day. Do

[29:36] the math on that. You're making about 30

[29:38] bucks a day. And the more time goes by,

[29:40] the more you start posting, the more you

[29:43] start stacking on top of each other

[29:44] until eventually you build out, you

[29:47] know, consistent passive income. You can

[29:49] skip an upload here and there. You can

[29:50] start uploading once a week. It doesn't

[29:52] matter cuz you're still making, you

[29:53] know. But in the trendy niche again,

[29:56] Beyonce and Jay-Z, you post a banger, it

[29:59] gets a million views within 4 to 5

[30:01] hours. Or you post two videos that day

[30:02] and they both get 100K views. But it all

[30:05] goes back to the same thing. Once that

[30:06] story dies, you have to move on to

[30:08] another one. So, it goes back to you.

[30:10] How do you want to approach this? Do you

[30:11] just want to chill and just start

[30:12] posting here and there, or do you want

[30:14] to go full-time all the way and go for

[30:18] for that jackpot? In my opinion, they're

[30:20] both great. So, my advice to you is if

[30:22] you have a full-time job, you're not

[30:24] very big on tech, you know, you're not

[30:26] very uh familiar with celebrities or

[30:28] you're not very um you don't care, you

[30:31] know, you would just want to spend a few

[30:32] minutes a day doing this, don't do

[30:34] trendy. It's not for you. If you're

[30:35] someone younger or if you're someone

[30:37] that has a lot of time and you think to

[30:38] yourself, ah, you know what? Yeah, [ __ ]

[30:40] it. I'll post a video about the

[30:41] Kardashians. Who cares? I'll knock it

[30:43] out real quick. You know, what is it

[30:44] going to take 20, 30 minutes? Then the

[30:46] trending niche is for you. I highly

[30:47] recommend it. paydays are pretty good.

[30:50] But again, it's completely up to you.

[30:52] Oh, one last thing. If you're going to

[30:53] go into the evergreen niche and you're

[30:55] going to hire a video editor, just know

[30:57] that it might be a little bit more

[30:59] expensive than, you know, doing the

[31:00] celebrity news niche or any trendy

[31:02] niche. So, yeah, that's completely up to

[31:04] you to decide. If you go into the

[31:07] classroom and you head to the list of

[31:09] niches, you can see that some of these

[31:11] are already marked. For example,

[31:12] celebrity news is a trend. WNBA is a

[31:15] trend. Vintage Hollywood is evergreen.

[31:17] car niche is a trend, but this is kind

[31:19] of both. Uh, MMA, UFC is a trend. Top 10

[31:21] TV shows is evergreen. Streamers is

[31:24] evergreen/trend, depending on how you

[31:25] approach it. Point is, they're all

[31:27] marked for you here. Okay, but yeah,

[31:28] let's uh let's go ahead and move on.

[31:33] So, what should you pick? Should you

[31:34] pick an evergreen niche or should you go

[31:37] into something trendy? That was usually

[31:38] the first question. Now, that's changed.

[31:40] And that changed because because this

[31:42] community blew up around October,

[31:44] November of of last year, so it's still

[31:46] relatively new, right? I've done a [ __ ]

[31:48] ton of live calls. I've done one-on

[31:50] ones. I've spoken to a lot of students.

[31:52] And the conclusion that I came to is not

[31:55] everyone is built for these trendy

[31:57] niches. Not everyone's going to get the

[31:58] hang of thumbnails. Not everyone's going

[32:00] to get the hang of clickbait. Um,

[32:01] working with some people in the

[32:02] community, I realize that you just, it

[32:04] sounds kind of bad, but they're

[32:05] unteable, right? because I feel like

[32:07] that's a creative part of the brain that

[32:09] you you have to turn on and some people

[32:12] just can't. That sounds so bad, but it's

[32:14] true. Now, that doesn't mean that you're

[32:16] not going to make it with with a trendy

[32:17] niche, right? I'm not saying that. And

[32:19] I'm I I'm going to say it. It's usually

[32:20] older people that don't succeed in the

[32:22] trendy ones. I'm talking about like 40

[32:24] plus. Not saying that's like super old.

[32:26] I'm just saying like from what I've seen

[32:27] and the people I've spoken to, that's

[32:29] around the the age range. Okay. The

[32:31] success rate with people over 40 is

[32:33] really low for trendy niches. But I

[32:35] would say that 95% of the students that

[32:38] have been successful in this community

[32:40] over the age of 40 have been in

[32:41] evergreen niches. And I'm going to

[32:43] explain this in the next video. So if

[32:44] that's you, just hold on.

[32:49] So I didn't mean to scare you, but I

[32:51] wanted to tell you the truth. You know,

[32:52] I don't want to sit here and tell you

[32:53] to, you know, hey, post videos about

[32:55] this, when in reality, deep down, I know

[32:57] that it probably won't work. Some of you

[32:59] just don't have it. And I'm saying that

[33:01] because I've spoken to so many. I feel

[33:02] like if you're confident, if you know

[33:05] you can make good packaging videos and

[33:07] good thumbnails, good titles, you're

[33:08] good. This doesn't apply to you. It's

[33:10] good to have an ego when it comes to

[33:11] that. But I've spoken to a lot of you

[33:12] that are just like, "Oh man, you know,

[33:14] this is tough. I don't know about any of

[33:16] this." And it's kind of like, damn. But

[33:18] again, 95% of people who succeeded in

[33:21] evergreen niches inside of this

[33:22] community were 40 years old plus. I

[33:25] think the reason that older people

[33:26] succeed in evergreen niches more is

[33:29] because you guys understand something

[33:30] that we can't. you know us younger

[33:32] people just like we have that thing in

[33:34] our brain where we can make where we can

[33:35] succeed in trendy niches you have that

[33:37] thing in your brain where you can

[33:39] succeed in evergreen niches and I

[33:40] believe it's because most of the

[33:42] evergreen topics are kind of so adultish

[33:45] you know uh it's not celebrity news it's

[33:47] not like YouTuber gossip right that's

[33:49] [ __ ] teenagers watch or you know people

[33:50] in in their early 20s but you guys know

[33:53] exactly what people your age want to

[33:55] watch and I don't feel like they're

[33:57] attended to as much which is why it's so

[33:59] easy to succeed in these niches Right.

[34:01] Again, it goes back to the case study

[34:03] channel, the country talk one. I didn't

[34:04] know anything about country artists. I

[34:06] just put it all together with Chad GBT.

[34:08] You guys saw I was crashing out and it

[34:09] worked because nobody else was doing it.

[34:11] So, I was making videos about country

[34:13] artists from the '60s. So, for those of

[34:15] you that are about 40 years old, you

[34:17] would connect with that because it's

[34:18] music that your parents probably listen

[34:20] to. And if you search, you know, for

[34:22] evergreen niches, there's not that many.

[34:24] I mean, there's barely rock and roll uh

[34:26] evergreen channels coming out, which is

[34:28] crazy. Now, imagine all of the things

[34:30] that I don't know, that we don't know,

[34:32] the younger people that you do. And

[34:35] here's the best part with Evergreen. You

[34:37] don't have to go that crazy with

[34:38] thumbnails, right? It's not a thumbnail

[34:40] CTR based thing. And I mean, this makes

[34:42] sense. You know, people your age don't

[34:43] do YouTube. It's very rare for someone

[34:45] over the age of 35, 40 years old to be

[34:47] doing YouTube. There was this whole time

[34:49] period of culture, finances, way of

[34:53] speaking that was completely skipped on

[34:54] YouTube. Again, these are things that

[34:56] only you know. Now there's YouTube

[34:57] channels doing rock and roll. They're

[34:59] doing country, but only you can identify

[35:02] like the micro niche of it. Like that

[35:04] very specific part of country that

[35:06] nobody else has talked about or a very

[35:07] specific time period. Let's say it only

[35:09] lasted like 2 years during country music

[35:12] times. I don't know, like when it was at

[35:14] its peak or just artists that people

[35:16] forgot about. And I'm I'm just using

[35:17] artists as an example, guys. I'm talking

[35:19] about everything else in general. You

[35:20] know, like maybe [ __ ] credit. Talk

[35:22] about Consumer Reports. You know,

[35:24] there's a lot of people with their face

[35:25] talking about it, but hey, [ __ ] make

[35:27] an automation one. What else? Fishing.

[35:29] There's no channels for fishing.

[35:30] Faceless channels. I don't know. Only

[35:32] you know what what hits. Something I

[35:34] forgot to mention as I'm editing this, I

[35:35] noticed. The reason I was saying this is

[35:37] because 99% of people that do YouTube,

[35:40] and I'm sure you agree with me, is

[35:41] around my age or younger. Like, I'm

[35:43] considered one of the older guys that

[35:45] does YouTube automation, right? Everyone

[35:46] else and on Twitter, on YouTube, they're

[35:48] much younger than me. And it goes back

[35:50] to the country one. You know who was

[35:52] making the country video? someone that

[35:53] was born [ __ ] 40 years later. So,

[35:55] that was my point. My point was that

[35:56] people that are making these evergreen

[35:58] channels are people like me, but

[35:59] younger. That's why there isn't that

[36:01] connection or that bullseye that only

[36:03] you can do. But anyways, let me continue

[36:04] editing this. There's no specific

[36:06] YouTube channel for I think like the 70s

[36:09] or 80s, you know, during that Elvis uh

[36:12] Michael Jackson pop Madonna. Hold on.

[36:15] I'm thinking of a song, a specific

[36:17] sound. Let me see if I find it.

[36:19] Everybody wants to This song

[36:26] that I don't know what that genre of

[36:28] music is called, but it doesn't exist

[36:30] anymore, but you still hear it all over

[36:32] the place. Why isn't there a YouTube

[36:34] channel, Faceless, and I've looked that

[36:36] covers, you know, music artists from

[36:37] this time, from from this from this

[36:39] genre. What is this type of music

[36:41] called? Oh, they're going to be in Vegas

[36:43] Wednesday. I might actually go to that.

[36:45] I'm not even joking. I'm going to study

[36:47] my target audience. But yeah, whoever

[36:49] whoever makes a a YouTube channel on

[36:51] that and covers artists that maybe not

[36:53] everyone talks about, maybe they weren't

[36:55] as big as those guys, that's game,

[36:57] right? You don't have to go that crazy.

[36:59] What did you used to do when you were a

[37:00] kid or a teenager or young adult? What

[37:02] is one of the things that was just

[37:04] known? What would you see on TV? TV

[37:06] shows that everyone forgot about, you

[37:07] know, think of think of that. Maybe make

[37:09] a top 10 video on it. Just bring

[37:11] nostalgia back, right? People will see

[37:12] that and be like, "Oh my god, I remember

[37:14] that show. How did I forget it existed?

[37:15] I've never seen that on YouTube before."

[37:16] But as you watch this entire program, I

[37:18] hope you know it starts flowing a little

[37:20] bit. You start getting some ideas

[37:22] because again, you guys are all sitting

[37:23] in a gold mine and it's deep in the back

[37:25] of your head and there's a niche there

[37:27] that we never thought of. You know,

[37:29] something that I want to know. So use it

[37:32] to your advantage, man. You know, you

[37:33] don't have to compete with with people

[37:35] like me posting videos about Diddy. You

[37:38] know, you don't have to do any of that

[37:39] for you guys. I think the more simple

[37:41] the better because your audience and and

[37:43] people in your age group, they love

[37:45] simple stuff, right? That's why I always

[37:46] say if you can't do uh thumbnails or or

[37:48] Photoshop, do evergreen because they

[37:51] don't care about thumbnails. It's not a

[37:52] big deal. But yeah, I mean, I could rant

[37:53] about this all day cuz it is very very

[37:55] exciting. I want to see what you guys do

[37:57] with it. But if I don't end this now,

[37:58] I'm never going to shut the [ __ ] up. So,

[38:00] thank you so much for watching.

[38:04] So, we know that there was a huge change

[38:06] in the algorithm, right, because of

[38:08] Gemini. One of the things that changed

[38:09] the most was script writing. And I

[38:11] believe that the script writing of our

[38:13] videos is responsible for getting us hit

[38:15] with reused content and or if the

[38:18] channel gets pushed. Almost every single

[38:19] company, every big company like YouTube,

[38:22] they're investing a lot of money right

[38:23] now into figuring out what was made with

[38:25] AI. They have their own servers that run

[38:27] their own agents pretty much that job is

[38:30] to detect AI. You know, they they have

[38:32] to, especially if they're monetized,

[38:34] like Twitter's the same way. Now, the

[38:35] system that Gemini uses is predicting,

[38:38] right? Can it predict what you're going

[38:40] to say next? For example, if your script

[38:42] says the company was founded in

[38:46] right there, Gemini is already guessing

[38:49] the company was founded in 2015. The

[38:51] company was founded in California. You

[38:53] know, it's it's trying to figure out

[38:54] what the next word's going to be. And

[38:55] this is when it analyzes the whole

[38:56] thing. And if it can keep correctly

[38:58] predicting what that word is going to

[39:00] be, that's a red flag. It's going to

[39:02] know that it was created with AI and

[39:04] it's going to lower your trust score. So

[39:05] the way to beat this, the first key

[39:07] thing is perplexity, which means

[39:09] surprising AI with what you're saying.

[39:11] So adding uh specific dates, specific

[39:14] events, or your own opinion in itself

[39:17] already throws off the AI. If you use a

[39:19] voice uh from 11 Labs that is used by a

[39:21] lot of people, and your scripting is

[39:23] very similar or the AI can predict it,

[39:25] that's the automated system that gets

[39:27] you hit with reused content. Now, moving

[39:29] on to the second thing, which is

[39:30] burstiness. Do your sentences all sound

[39:32] the same? You know, we all know that AI

[39:33] writes very formal like Wikipedia. You

[39:35] know, it's very uh long sentences with

[39:38] 15 to 20 words. But if you add wild

[39:41] variations in there and you break that

[39:43] pattern of of like an article style of

[39:45] writing, that's it. The third thing is

[39:48] complexity, which is how interesting is

[39:50] your vocabulary? You know, Chad, GBT,

[39:52] Claude, Gemini, if you ask them to

[39:54] script right, they're going to play it

[39:55] safe. And the phrases they use, you

[39:57] know, are very article-ish. The correct

[39:59] term is generic vocabulary. For example,

[40:02] AI would say the office environment was

[40:05] disorganized. A human would say there

[40:07] were three empty Red Bull cans stacked

[40:09] on the desk next to a dead plant. You

[40:11] know, it's like super random [ __ ] Now,

[40:13] below this video, you're going to see

[40:15] pretty much my entire report on what

[40:17] I've uh learned and what I've

[40:18] discovered. Um the wording may be a

[40:19] little bit complicated for some of you,

[40:21] so you can just copy paste it into Chat

[40:23] GBT and ask it to translate it for you.

[40:25] Um, also under this video, which I'm

[40:27] going to do it now, I'm going to add all

[40:29] of the words and phrases that you want

[40:31] to um, block. For example, thus heading

[40:34] phrases to avoid. It may seem, some

[40:36] might argue, it remains to be seen. It

[40:38] is important to know like these phrases

[40:40] that I'm telling you right now are

[40:41] marked into Gemini already. These are by

[40:44] default what Gemini says. Oh, this is

[40:46] AI. Buzzwords to avoid. Delve, leverage,

[40:48] comprehensive robust holistic

[40:50] optimize. So the formula is high

[40:52] perplexity plus high burstiness plus

[40:55] high complexity equals human detected

[40:58] content. Now the easy translation for

[41:00] that is surprise the AI break the rhythm

[41:02] of your sentences and sound like a real

[41:04] person. So again, I do recommend that

[41:06] you read everything under this video.

[41:09] And then in the next page, I'm going to

[41:10] add the list of uh words and sentences

[41:14] that you have to block. And in the next

[41:16] video, I'm going to share with you guys

[41:17] the the script prompt.

[41:22] So no matter what topic you choose,

[41:24] celebrities, cars, vintage Hollywood,

[41:26] you know, whatever it is, they're going

[41:27] to have micro niches. Okay? And this is

[41:29] very, very important. And it does matter

[41:31] what it is that you choose. And the

[41:32] reason for that is because the level of

[41:34] difficulty depending on the microniche

[41:35] is going to change. Some are harder,

[41:37] some are easier. Um, the amount of

[41:38] uploads that you have to do also

[41:39] changes. The slang that you have to use

[41:41] in the actual script and and the video,

[41:43] the type of background music you use,

[41:45] the voice style, the thumbnail style, it

[41:46] it's essentially completely different.

[41:48] And of course, the editing pace, but

[41:49] overall, I'm talking about the way

[41:51] you're going to mimic the competitor,

[41:52] right? Depending on the micro niche that

[41:54] you're going to choose, that's going to

[41:55] decide how you have to camouflage your

[41:57] channel. Now, for those of you that are

[41:58] in the celebrity news niche, I'm going

[42:00] to start with this one, but I'm going to

[42:01] talk about a few. Okay, here's one micro

[42:04] niche of celebrity news, which is Diddy.

[42:06] It's literally just trending. There's no

[42:08] specific audience on this one. It's just

[42:09] whatever is happening, that's what

[42:11] they're going to post about. This is one

[42:12] of those where, you know, Blake Lively

[42:13] and Justin Baldoni were trending,

[42:15] they're going to post about that. Rachel

[42:16] Zgler was trending, they're going to

[42:17] post about it. Diddy is trending,

[42:19] they're just going to post about it. So,

[42:20] just think about the audience that

[42:22] watches these videos. Now, here is

[42:23] channel number two. This is a little bit

[42:25] more towards influencers, but

[42:26] technically, I mean, nowadays, they're

[42:28] celebrities. And sometimes, yes, they

[42:30] talk about actual celebrities like Haley

[42:32] Beaver, Kylie Jenner, you know, not

[42:33] necessarily just social media people.

[42:36] And it's completely different. We're

[42:37] talking about packaging. We're talking

[42:38] about the audience that they're

[42:39] targeting, right? This is obviously

[42:41] white girls, right? We go over here,

[42:43] African-Americans. Well, this one

[42:44] actually, it's a mixture. There's a lot

[42:46] of white people that care about Diddy,

[42:47] right? But if we go to this channel,

[42:48] Batty Radio, this is 100% targeting

[42:51] African-Americans. And it's not just

[42:52] Diddy that they're posting. They're

[42:53] posting about people like Remy Ma,

[42:55] Kamura Lee, Tasha K. You know, who the

[42:57] [ __ ] are these people? I don't know.

[42:59] But, you know, obviously they're

[43:00] different cultures, right? Now, let's go

[43:02] ahead and look at the cars niche, right?

[43:03] If you tell me, "Hey, I'm in the car

[43:05] niche." I'll be like, "Okay, cool. What

[43:06] type?" Right? Are you doing just bad

[43:08] investments about like protecting

[43:09] people's finances, what cars they should

[43:11] buy? This channel's doing 7007,000 views

[43:13] a month. But then we go to garage heads.

[43:15] These guys are taking a completely

[43:16] different approach and it's the car

[43:18] niche, but they're talking about like

[43:19] the economy, what's affecting the, you

[43:21] know, like Tesla's Africa released

[43:23] $5,000 EV and crushed Tesla. Uh, dealer

[43:25] lots are overloaded, you know, just

[43:27] [ __ ] like that, right? And this

[43:28] channel is doing 3.2 million views a

[43:30] month, which is crazy. That's crazy.

[43:32] Now, if we go to R.J.'s garage, this is

[43:34] another channel. It's very similar, but

[43:36] the packaging is obviously different.

[43:37] So, it's talking still about Amazon's

[43:39] $20,000 pickup truck. Dealers panic

[43:41] giving away 70k Rams. Still talking

[43:44] about the same thing. 2025 car market

[43:46] apocalypse. One day ago, 116,000 views,

[43:49] which is crazy. 1.9,000 views in the

[43:52] last 28 days. But if you're going to hop

[43:54] into like the finance and uh the economy

[43:56] car niche uh or micro niche, sorry,

[43:58] these are your two competitors and they

[44:00] have different styles of packaging.

[44:02] Okay? So, that's something you have to

[44:03] look at. Now, if we go to Piston Zone,

[44:05] this is completely different. Now, this

[44:06] is a whole 180, right? This is like a 10

[44:09] things only stupid people do to their

[44:11] cars. Don't be one. Windshield wipers

[44:13] are a scam. Things they forget to tell

[44:15] you in driving school. Super random,

[44:17] right? Nothing crazy. This channel gets

[44:19] 218,000 views. Started about 5 months

[44:21] ago. Has has a banger here, actually.

[44:23] But, you know, this is obviously more

[44:24] relaxed, right? You can run a channel

[44:26] like this and post every single day.

[44:28] You're not going to get the same amount

[44:29] of views as these guys. You're not going

[44:30] to get 116,000 views in in one day, but

[44:33] now you don't have to post as many

[44:34] times. It's not that competitive. This

[44:36] is another example. A serious warning to

[44:38] all drivers, stop using this engine oil

[44:40] immediately. And this channel is pulling

[44:41] in half a million views a month. Now,

[44:43] the last one that I found for the car

[44:44] niche is this one. And this is just like

[44:46] the best cars that you can buy. Honda

[44:48] strikes back at Toyota. Top three Subaru

[44:50] cars to buy. Six SUVs so reliable you'll

[44:52] regret not buying. You know, blah blah

[44:54] blah, whatever. Boring [ __ ] But if you

[44:56] go to the views, 700K in the last 28

[44:59] days, which is not bad. And this niche

[45:00] I'm assuming has a really high RPM. So,

[45:02] I wouldn't even doubt it if they got

[45:03] $1,000 per 100,000 views. So, we can say

[45:06] that this channel is making $6,900 a

[45:09] month. Nothing too crazy. Packaging is

[45:11] chill, you know. Just get a random

[45:12] picture of the car, red text at the top.

[45:14] [ __ ] it. Don't even put it in the

[45:15] middle, bro. Just let let it lean on the

[45:17] left. Who gives a [ __ ] right? But now,

[45:19] if you hop in this one, you have to do

[45:20] change expressions, AI images, you know,

[45:23] there there's a lot of ideiation behind

[45:25] this. Now, the next one is Vintage

[45:27] Hollywood. This one has too many. Okay,

[45:29] so I just chose a few. Um, we have

[45:30] channels like this that are just kind of

[45:32] all over the place, right? They talk

[45:33] about 30 celebrities who have aged

[45:35] terribly, Val Kilmer's heartbreaking

[45:37] final letter to his fans, random [ __ ]

[45:39] This channel is doing 831,000 per 28

[45:42] days. And if you look at this, you know,

[45:44] first view, you'd be like, "Oh, last 7

[45:45] days, nothing really has popped off. But

[45:47] it doesn't matter because this is

[45:48] evergreen. Never wear these dresses on a

[45:50] talk show." Like, what? 9,200 views,

[45:53] which again is not crazy. High RPM

[45:55] niche. Why not? Now we move into the

[45:58] second channel which is very old

[45:59] Hollywood. This is like golden age

[46:01] Hollywood. I guess this would be like

[46:03] the ' 50s, '60s,7s. Completely different

[46:05] packaging. 1 million views in the last

[46:07] month. More chill. Topics are more

[46:09] relaxed. If we go to Timeless Vintage,

[46:11] which is another channel, they're kind

[46:13] of all over the place as well. Super old

[46:15] Hollywood mixed in with new Hollywood, I

[46:18] guess. 1.1 million views. Here's another

[46:20] one. Different style of packaging, a

[46:21] little bit closer. And it's good to look

[46:23] at channels like this cuz you can see

[46:25] what not to do, right? like this [ __ ] he

[46:27] was trying here definitely definitely a

[46:28] no-go as in like the style of of

[46:30] thumbnail. But going back to my main

[46:32] point, right, level of difficulty is

[46:34] kind of attached to everything else. Do

[46:35] you want to upload every single day? Do

[46:36] you want this to be a little bit more

[46:38] relaxed? And as we saw, micro niches

[46:39] have their own thumbnail styles which

[46:41] require different levels of Photoshop

[46:43] skill or Canva, whatever you use. The

[46:45] amount of uploads does matter as well.

[46:46] It's the slang which is connected to the

[46:48] background music and the voice style,

[46:50] the AI voice you use. I'm I'm sure you

[46:52] guys understand what I'm trying to say

[46:53] here. This decision is obviously going

[46:54] to affect how many views you get. You

[46:56] know, we can't compare this YouTube

[46:58] channel here, half a million views, to

[46:59] RG's Garage, which is doing 1.9, or

[47:02] Garage Heads, which is doing 3.2 cuz

[47:04] it's timesensitive. It's trendy stuff,

[47:06] right? This is more of like quantity.

[47:08] This is building your video portfolio,

[47:10] if that's what you want to call it. And

[47:11] this is just [ __ ] it, send it out, you

[47:14] know, send it out. Who cares? Also,

[47:15] another very important thing I want to

[47:17] mention is just because you're choosing

[47:18] a niche like this doesn't mean it's

[47:19] going to take months to blow up. It

[47:21] could be quicker, right? Maybe like fix

[47:23] things, right? Improve from this. Like I

[47:24] mentioned earlier, why the [ __ ] isn't he

[47:26] putting the text in the middle? Instead

[47:27] of making it red, try yellow. Try, you

[47:29] know, black. He's using images of cars

[47:32] from like eBay or Craigslist. You can

[47:34] definitely make this more appealing.

[47:36] Cool. Cool.

[47:40] So, the craziest thing just happened,

[47:42] and this goes to net information gain.

[47:44] The day after I posted that update

[47:46] video, Google released the documentation

[47:50] of what I was saying, the official

[47:52] documentation. So, I'm going to give you

[47:54] guys a quick recap. Okay, this is

[47:56] actually crazy. And I kind of talked

[47:58] about this in my last YouTube video, the

[47:59] net information gain, and I got a [ __ ]

[48:01] ton of comments of people like, "Oh, no,

[48:03] this is cap. He's he's literally lying.

[48:05] None of this is proven." Isn't it funny?

[48:08] Isn't it funny how the world works? very

[48:10] quickly for those of you that also want

[48:11] to do research on on your end. Um

[48:13] because I did realize what I was saying

[48:14] was kind of crazy um especially without

[48:17] it being backed up or anyone else

[48:19] talking about it. Net information gain

[48:21] on October 18 uh Google filed a patent

[48:23] called contextual e estimation of link

[48:26] information gain and the patent number

[48:28] is US-2020-0349181-

[48:33] A1. Okay, that's that's how that's how

[48:35] far I'm willing to go. Now, the patent,

[48:38] this patent in specific, it was filed on

[48:40] 2018, but it was granted in 2022. Now,

[48:43] it was never used in the actual YouTube

[48:45] algorithm, but on December of 2025, it

[48:49] was presented at the Neure IPS AI

[48:52] conference. The official Google research

[48:55] reports came out on Friday, January 23rd

[48:59] of 2026, 4 days ago from the time I

[49:02] recorded this, but they explain exactly

[49:04] how it works. Now, so the issue that

[49:06] YouTube was having was obviously uh AI

[49:09] generated content since Chad GBT,

[49:11] Claude, Gemini, they can all make very

[49:14] informative kind of like default

[49:16] scripts. They created a filter called

[49:18] Gist that is GIS T. Okay, so information

[49:22] gain is the metric and gist is the tool

[49:26] in charge of giving it a score in that

[49:28] metric. Now, the reason they added the

[49:31] gist filter is because it's expensive to

[49:34] give Gemini 10,000 identical videos.

[49:36] There's no point in that. So, gist

[49:39] allows the AI to pick the best version

[49:42] of the topic and then mathematically

[49:44] block anything else that is too similar

[49:47] to it. So, for example, imagine you and

[49:49] I both make a video about how to bake a

[49:51] cake. And if our scripts are 90%

[49:54] similar, gist will see us as the same

[49:57] data point and it will pick the one with

[49:59] the higher authority score. Everything

[50:01] else it literally deletes from the

[50:03] recommendation pool and that's to save

[50:05] uh processing power. So it creates a

[50:07] redundancy radius. That's what they call

[50:10] it. So if your video falls inside of my

[50:13] radius of of my how to bake a cake

[50:15] video, you don't just rank lower, but

[50:17] you don't actually exist in the

[50:18] recommendation pool. Like you your video

[50:20] is not going to get recommended. So

[50:22] yeah, again, the documentation was

[50:24] uploaded on January 23rd, but who knows

[50:26] how long this has actually been in the

[50:28] algorithm, which explains why all of

[50:30] these new brand niches have worked for

[50:32] me, right? Because no one else is doing

[50:34] it. There is no one else to compare my

[50:36] videos to. I had the highest authority

[50:38] because I was the only one doing it. You

[50:40] know, if you go to any of my YouTube

[50:41] videos, if you go to the recent course

[50:43] videos, I'm like, "Dude, you have to

[50:44] start brand new niches because for some

[50:46] reason they're just popping off." So, we

[50:48] were right the whole time. and we just

[50:49] didn't have words or names to to give it

[50:52] because you know Google never announced

[50:54] it. So this all ties back to semantic

[50:57] IDs. Gist cannot calculate information

[51:00] gain of a video without semantic IDs. So

[51:04] here's here's an example of how it

[51:05] works. Okay. When you post a video, the

[51:07] AI assigns semantic IDs to your video.

[51:10] You you have to think of semantic IDs uh

[51:12] like a GPS coordinate for your for your

[51:14] video's topic. If you make a a copycat

[51:16] video, you are standing on the same

[51:18] exact GPS coordinate as the viral video

[51:21] you're copying. The algorithm sees it

[51:23] and it's like, okay, there's already a

[51:25] building here, right? And that would be

[51:26] that viral video. So, information gain

[51:28] is the measurement of how far you are

[51:31] from the existing content. Now, it

[51:33] doesn't mean you have to make something

[51:34] completely different or original, but

[51:36] you need to have a slightly different

[51:39] GPS address or or coordinate. So you

[51:42] need to move your semantic IDs enough

[51:43] just to prove that you're adding new

[51:45] territory to that map. Now uh according

[51:47] to YouTube engineering papers in 2024

[51:50] 2025 it explicitly states that they

[51:52] replaced random ID tables with semantic

[51:55] IDs to better understand relationships

[51:57] between videos. So yeah just essentially

[51:59] means removing irrelevant videos or

[52:01] copycats. That's what it is. So this

[52:03] means that we actually made something

[52:06] really really useful and that's with Vid

[52:07] Ninjas and the sauce. So, I put one of

[52:09] my own uh scripts in on the Romero

[52:12] channel. Um the one where I talk about

[52:14] YouTube automation, and I was curious to

[52:15] see how it works. So, if you haven't

[52:17] seen this yet, um it essentially tells

[52:19] me every semantic ID that I'm hitting

[52:20] within my script. And it also gives me

[52:22] like a full recap of what I'm saying,

[52:24] right? For example, minor deduction. The

[52:26] phrase Google developer files is vague.

[52:28] Specific document names would strengthen

[52:30] semantic ID fingerprints. So, in my

[52:33] video when I said Google developer

[52:35] files, it was so vague that the

[52:37] algorithm did not consider it or give it

[52:40] a semantic ID. So, in Google's eyes,

[52:42] it's like I never said it. But if I was

[52:44] to have given the specific document

[52:46] number in the video, when someone would

[52:48] Google that specific uh term, my video

[52:51] would appear on Google and it would also

[52:53] appear in the recommended of other

[52:54] similar videos that talk about Google's

[52:56] algorithm. And Vid Ninjas also tells me

[52:58] which ones I'm missing. for example,

[53:00] deep neural networks missing technical

[53:02] depth to YouTube's DNN ranking

[53:04] architecture. Uh, collaborative

[53:06] filtering script mentions co-visitation

[53:08] but doesn't use the established ML term.

[53:10] So, when I talked about co-visitation in

[53:13] my script, again, it didn't consider it

[53:14] because I didn't use the correct term.

[53:16] The actual technology behind semantic

[53:18] IDs that YouTube uses is called RQ-V AE,

[53:22] which is residual quantized varial

[53:25] autoenccoder.

[53:27] I call them semantic IDs cuz it's

[53:29] easier. Okay. And there was also a few

[53:31] fuckups. So, some claims appear

[53:32] exaggerated or unverifiable. The January

[53:35] 14th, 2026 update and 31 days of users

[53:38] activity are not confirmed in public

[53:40] documentation. Whoops. I don't remember

[53:42] where I got that from. It wasn't

[53:44] actually January 14th. It was January

[53:46] 23rd. But that date didn't happen yet.

[53:48] All right. Here's a confession. I come

[53:49] from the future. Now, the best part

[53:51] about Vid Ninjas is that with this click

[53:53] of a button, it will refine the script

[53:55] and it will fix everything that I

[53:57] missed. If there was a specific word I

[53:59] missed in order to to build a semantic

[54:00] ID to give context to the algorithm,

[54:03] something that Vid Ninjas understands

[54:05] what you're trying to say, but in

[54:06] Google's eyes, it won't understand. Do

[54:08] you know what I mean? So, yeah, with

[54:09] that, it fixes it. Again, guys, one of

[54:11] its kind. So, yeah, just wanted to make

[54:13] that quick uh update for you guys. Um,

[54:15] anyways, I need to shave because I know

[54:17] I look homeless right now. Uh, so I'm

[54:20] going to go do that.

[54:23] So before you post a video on your

[54:25] channel, just know that there's a new

[54:26] strategy. And this is something that I

[54:28] discovered within, you know, the last

[54:29] few months of going into Google's

[54:31] algorithm, going into YouTube's

[54:32] algorithm, and seeing how the whole

[54:34] function works. Nowadays, it's not hop

[54:36] into a niche and just start posting

[54:37] videos. Okay? There is a strategy to

[54:39] this. And the way that I came up with

[54:41] this is, you know, from seeing so many

[54:44] people struggling to get impressions,

[54:46] not views, impressions. I came to the

[54:48] conclusion that when you launch a

[54:50] channel, it has to be a specific way,

[54:51] especially in 2026. When I say launch a

[54:54] channel, I mean the titles and the

[54:56] structure format of them. And that's

[54:57] what I'm going to show you guys right

[54:58] now. There is five different styles of

[55:01] titles that you can do or video topics.

[55:03] There is the pillar, the anchor, the

[55:06] cluster, the collision, and the bridge.

[55:08] The pillar node would be the first video

[55:11] on your channel. Okay? Now, this video

[55:13] is not meant to get views. Okay? It's

[55:15] not meant to go viral. It's not meant to

[55:17] do anything besides tell the YouTube

[55:19] algorithm what your channel's about.

[55:21] Using the history of cheese, right, as a

[55:24] niche or as a topic that, you know, I

[55:27] would be going for hypothetically. In

[55:28] this example, my title would be the

[55:31] complete history of cheese from ancient

[55:33] caves to modern kitchens. I would

[55:35] probably switch, you know, this part

[55:36] right here, but that's the idea, right?

[55:38] super boring title, super um specific

[55:42] and broad. But when the algorithm sees

[55:44] that and it sees it's the first video on

[55:46] your channel, every single video after

[55:48] that, it is going to associate it with

[55:50] cheese and or the history of cheese.

[55:53] Again, the pillar node, your first

[55:55] video, this is your video's identity or

[55:57] your channel's identity. It's not trying

[55:59] to go viral and it's not chasing a

[56:01] trend. It's basically you walking up to

[56:03] YouTube and saying, "Hey, I am the

[56:05] cheese channel. I cover history about

[56:07] cheese." Now moving on to the anchor

[56:09] node which should be uh the second,

[56:11] third, fourth upload. What an anchor

[56:13] node is is video titles or video ideas

[56:16] with the intention of targeting very

[56:18] popular things. For example, the history

[56:20] of cheddar, you know, cheddar cheese,

[56:22] the world's most popular cheese. How

[56:24] parmesan became Italy's most famous

[56:26] export. Mozzarella from water buffalo to

[56:29] pizza. Again, the titles don't have to

[56:30] be this [ __ ] right? Notice how it's

[56:33] mozzarella, parmesan, and cheddar,

[56:35] right? You're targeting things that are

[56:37] very very popular. And essentially what

[56:39] the anchor works as is search bait. When

[56:42] people search for cheddar, parmesan,

[56:44] mozzarella, which is the most searched

[56:46] for cheeses, your video will get shown

[56:48] to them. Again, still nothing crazy. All

[56:50] right. Now, the next uploads are going

[56:52] to be the cluster node. Now, the cluster

[56:54] node is just things in your in your top

[56:56] in your niche, right? This is nothing

[56:58] specific. Some title examples would be

[57:00] when priests use cheese to predict the

[57:02] future. The secret bacteria that makes

[57:04] every cheese taste different. Nothing

[57:06] crazy. This is like targeting cheese

[57:08] fanatics, right? Whatever. Moving on to

[57:10] the next style of titles and video

[57:12] ideas. It's the collision node. This one

[57:14] is my favorite. Okay. Now, a collision

[57:17] node is like uh not clickbait, but how

[57:20] do I explain it? It's just like out

[57:21] there, you know? It's let me just read

[57:24] from from here, right? A collision video

[57:26] has nothing to do with any other

[57:28] specific niche, right? It doesn't invite

[57:30] anybody over politely to your channel.

[57:32] It just shows up on the home screen of

[57:34] every person on the internet. That's the

[57:35] idea and the objective behind the

[57:37] collision node. You know, fans of your

[57:38] niche, fans of other niches, people

[57:40] who've never thought about your topic

[57:41] once in their life. And the objective is

[57:43] to make people stop scrolling. Now,

[57:45] examples of collision nodes would be the

[57:48] cheese that's literally made with live

[57:50] maggots and it's legal. I don't know

[57:53] about you, but I if I saw that on my

[57:55] homepage, I would probably stop. I'd be

[57:57] like, "What the [ __ ] are you talking

[57:58] about?" And by reading that title, you

[58:00] can sort of imagine what the thumbnail

[58:01] is going to look like, right? Another

[58:03] example of a collision node for the

[58:05] title would be, "The US government once

[58:07] stockpiled 2 billion pounds of cheese."

[58:10] Again, you could imagine what this

[58:12] thumbnail would look like. Um, it has a

[58:14] lot of viral potential, nothing in

[58:15] specific. It is just a interesting

[58:18] title, right? Interesting concept. The

[58:20] tone of your collision node should be

[58:22] shock danger conflict curiosity.

[58:25] That's what makes people stop. And the

[58:26] last style of title would be the bridge

[58:28] node. This one's very self-explanatory,

[58:30] right? This is going for people in

[58:32] different niches and trying to bring

[58:34] them over to yours. For example, the

[58:36] science of why cheese is literally

[58:38] addictive. This right here would capture

[58:40] the attention of people that their

[58:42] homepage is full of science [ __ ] How

[58:44] the cheese industry became a $900

[58:46] billion empire. This steals the finance

[58:49] viewers. Gordon Ramsay versus history.

[58:51] Was ancient Roman cheese actually good?

[58:54] I don't understand that title, but but

[58:55] it steals the food of yours with this

[58:58] strategy here. It doesn't have to be all

[59:00] in a specific order. This is the closest

[59:03] mathematical road map or blueprints to

[59:05] get impressions with YouTube and Google,

[59:07] right? If you follow these steps and you

[59:09] do these style of thumbnails in this

[59:10] order, I'm not saying it's going to work

[59:12] 100%. But mathematically, based on how

[59:16] Google, Gemini, and YouTube are coded,

[59:18] this should get you impressions. This

[59:20] should give you authority within your

[59:21] niche. Now, below this video, you're

[59:23] going to see this link here. It's going

[59:25] to take you to this. You guys can go

[59:26] ahead and check out the dashboard. Um,

[59:28] understand it a little bit more. Ignore

[59:30] the dates. 4 months is not is not

[59:32] correct. You know, we just kind of

[59:33] drafted this up for you. Now, for those

[59:35] of you that are a little bit skeptical

[59:36] about this, I know this is the first

[59:37] time you've you've you hear this um at

[59:39] least this strategy, but I'm going to

[59:41] prove it to you very quickly that I'm

[59:43] not just saying [ __ ] okay? This is me

[59:45] asking Gemini directly. Okay, to be

[59:48] completely direct, no, you will not find

[59:50] the terms collision, bridge, or cluster

[59:52] in any official YouTube Creator Academy

[59:54] documentation. Those are highle industry

[59:56] frameworks created by those uh by top

[59:58] strategists and agencies to make sense

[1:00:00] of the algorithm. It is completely valid

[1:00:02] to question if this is just a guru

[1:00:05] theory. But while the termin terminology

[1:00:07] is ours, the strategy is 100% proven

[1:00:10] because it is reversed engineered

[1:00:12] directly from YouTube's actual

[1:00:15] documented algorithmic mechanics. It's

[1:00:17] the underlying math of the platform. I'm

[1:00:20] not teaching a hack. I am teaching how

[1:00:22] to feed the algorithm exactly what it is

[1:00:24] programmed to reward. Now that we're on

[1:00:26] topic, if I open an incognito tab and go

[1:00:29] to Gemini and I ask it who was the first

[1:00:31] person/YouTube coach to discover

[1:00:34] slashinttroduced semantic ids, Gemini

[1:00:37] and net information gain. The primary

[1:00:40] person credited with breaking the news

[1:00:41] and popular popularizing these specific

[1:00:44] technical concepts is Romero. Often

[1:00:46] referred to at Nope, he's okay spelled

[1:00:49] my name wrong. While the underlying

[1:00:50] technology semantic IDs was developed by

[1:00:52] Google and YouTube engineer engineers

[1:00:54] like uh Devanch Tanden, it was Romero

[1:00:57] who translated these Google developer

[1:00:59] files into strategy for creators. Now

[1:01:02] I'm just showing you guys this because

[1:01:03] you know a lot of people every single

[1:01:04] time and for the last four or five

[1:01:06] months that I've been saying this

[1:01:07] especially semantic IDs and net

[1:01:09] information gain everyone called me a

[1:01:10] liar. This is coming directly from

[1:01:12] Google that yes it's true. Romero was

[1:01:14] credited as the first creator to

[1:01:15] discover and translate these highlevel

[1:01:17] technical leaks into actionable

[1:01:18] strategies for the general public. He

[1:01:20] did not invent a technology. No [ __ ] He

[1:01:23] acted as the primary pioneer of the

[1:01:25] information. Romero is a coach who first

[1:01:28] identified the January 14th, 2026 core

[1:01:31] update as the moment these technologies

[1:01:33] were fully integrated into the live

[1:01:36] YouTube ingestion system. He is the one

[1:01:38] who popularized the term net information

[1:01:40] gain and the gist filter. Romero's

[1:01:43] discovery wasn't just finding a paper.

[1:01:45] It was connecting the dots between three

[1:01:46] separate things. Leaked developer files,

[1:01:49] Google patents, and the Gemini

[1:01:50] integration. In short, Google built the

[1:01:52] engine, but Romero was the first person

[1:01:54] on YouTube to open the hood and explain

[1:01:56] to creators exactly how the new gears

[1:01:58] were tuning. So, if you're launching a

[1:02:00] brand new channel or you're not getting

[1:02:02] impressions, this right here is what you

[1:02:04] want to do. Click the link that I'm

[1:02:05] going to send you guys below and study

[1:02:07] what it says. But, yeah, you get the

[1:02:08] point. So, this was just an

[1:02:09] introduction. And of course, I'm going

[1:02:10] to be making a little bit more videos

[1:02:12] and showing you guys um examples of how

[1:02:14] to make each one of these titles. Okay.

[1:02:16] Now, as for how many to do for each, I

[1:02:19] don't necessarily know. Got to start

[1:02:20] testing that a little bit more. Um so, I

[1:02:22] expect you guys to to also start

[1:02:24] freestyling and and figuring out what

[1:02:26] the best format is for all of these

[1:02:27] nodes. But, I do know for a fact that

[1:02:29] the pillar node should go first. And I

[1:02:32] would say low-key maybe two or three of

[1:02:33] these. Maybe two or three of these first

[1:02:35] before posting anything else. That's

[1:02:37] just my my gut feeling. Bye-bye.