Full transcript
0:00Jachi BT's latest model, Astra, just
0:02solved the hardest part of YouTube,
0:04especially for business channels. So,
0:06what do you think the hardest part of
0:07YouTube is? Is it the scripting or the
0:10editing or getting really good on
0:11camera? Actually, the thing that trips
0:13up my students more than anything else
0:15is just picking the right topics and
0:18titles that their customers care enough
0:21to watch. That is the part of YouTube
0:22that's not only the hardest, but it's
0:25the part that if you can get right,
0:27you're like 80% of the way there. and
0:29Chat GBT6, otherwise known as Astra, it
0:32came out last week and right away it was
0:34blowing past every industry benchmark,
0:36even outdoing Claude Fable in terms of
0:39logic tasks. So, I was like, I wonder if
0:41this can just take over this part of
0:42YouTube for people and tell them exactly
0:45what types of videos they should be
0:47making that are just going to guarantee
0:49them lots of views from the right
0:50viewers. So, I tried it and honestly, I
0:53am kind of blown away by how good it is.
0:55I really think we're there now. So, in
0:58the rest of this video, I'm just going
0:59to show you how you can use ChatgBT
1:01Astra's really big brain to work on your
1:04channel, whether or not you've even
1:06started on YouTube yet. It works either
1:08way. And this works especially well if
1:10you're the kind of channel where you're
1:12really niched down and there might not
1:14be a lot of search volume for what
1:16you're trying to make videos about. And
1:17even if there is search volume, you
1:20probably have a lot of competition out
1:22there. So, this really is just going to
1:23help you kind of find your openings and
1:25your opportunities. So, I want to talk
1:27really quick about what Astra is and
1:30like what people are saying about it.
1:31So, basically, people are saying that it
1:33is the most agentic AI we've had yet.
1:36Meaning, you can just give it a goal
1:38rather than it like telling it exactly
1:40what to do. You give it a goal and then
1:43it's going to work out how to get there
1:45on its own. And people are also going so
1:48far as to say prompt engineering is now
1:50dead, right? You can just tell it what
1:52you want it to get to and it's going to
1:54figure all that out. So, you don't
1:56really have to be a master prompter
1:58anymore to get it to work. So, that's a
2:01pretty big claim and we will get to that
2:03a little later on. And I do want to
2:04address something before we go any
2:06further. So, whenever I do a video like
2:07this that kind of talks about using AI
2:10to come up with topics or titles or help
2:12you out in any way, I always get the
2:14comments from people saying, "Well,
2:15isn't this just a race to the bottom?"
2:17Right? If AI is helping us do all this
2:19stuff, isn't it at all just going to
2:22feel the same? Is it going to be AI
2:23slop? Here's the deal. You always have
2:26to do research anyway, right? If you're
2:29going to be starting on YouTube, the
2:31last thing you want to do is just go on
2:33feelings and like what you want to make
2:36videos about, that's a recipe for
2:38disaster. And that's why most channels
2:40never really make it because they're
2:42making the wrong videos. So, as long as
2:44we have to do research anyway, you can
2:46either do it with your own head and you
2:48can agonize over it for a long time. And
2:51trust me, I've seen a lot of people do
2:52it that way and they kind of always
2:55fumble the ball. So, this is just a much
2:58more scientific way. It actually
2:59crunches the data. AI is way better at
3:03uh coming at seeing patterns than we as
3:05human beings are, especially now with
3:08Astra. So, is it a race to the bottom?
3:10Not even close. This is your chance to
3:12kind of rise up to the top if you do it
3:14right. Okay. Okay. So, what I'm planning
3:15on doing in this video is I'm going to
3:17put this to the test on two different
3:19channels that you guys submitted. So,
3:21one is a channel that's been struggling
3:23for about 12 years, never got any
3:24traction, and the other one is just a
3:26brand new channel. I think there's a
3:28couple videos from like 3 years ago, but
3:30it's basically a clean slate. So, that's
3:31the setup. Now, we need to understand
3:33how this is going to actually work in
3:35practice. So, think of ChatGBT as the
3:38one who crunches all the data, right? It
3:40finds the patterns. It knows what to
3:42look for. It gives you your suggestions,
3:44but it has to get its data from
3:47somewhere. It can't just make it up.
3:48That's that would be the worst thing
3:50ever. So, what we have to do is actually
3:52connect it to another tool called Vid
3:54IQ. Now, Vid IQ is actually connected
3:56into the back end of YouTube. So, it can
3:59see if you have a channel already, it
4:00can see all of your numbers. It can see
4:02most of your competitor's numbers, their
4:04titles, what's in the videos. It can see
4:07all the transcripts. So, this is where
4:08the data comes from. This is how it's
4:10going to know what is working, what's
4:13not working. So, and I did work out a
4:15deal for you guys. You can start today
4:16for a dollar for your first month. So,
4:18you can try it. You can cancel it after
4:20that if you want to. But once you're set
4:22up inside of there, what you need to do
4:24is get the connector that you're going
4:25to use to connect to ChatGpt. It's
4:27really easy, I promise. You're just
4:29going to go down to your name down here,
4:31and you're going to go up to MCP.
4:34You're going to choose ChatGpt, and then
4:36just make sure you're logged into
4:37ChatGpt. And if you are, you're just
4:38going to click this button right here.
4:40And then it make sure you're all set up
4:42on chatbt. It's a one-click install. So
4:44then I'm just going to click try in
4:46chat. And the first thing we want to do
4:49is we want to make sure that we are
4:51choosing the right model. We want to be
4:53using Astra here. So uh we're going to
4:56choose GPT6 Astra. And then you can
4:59choose your own effort level. Um, I find
5:03that the ultra at the very top that does
5:06run through the credits the fastest. So
5:08maybe you want to back it down to max,
5:09but if you've got the credits and you
5:11don't mind spending it, it might work a
5:13little better for you on the far end
5:14here. All right, so our first case study
5:16is a guy named Delvirus. He is a tattoo
5:18artist. Um, so I would just go here and
5:21I would click on the little microphone.
5:23It's a lot easier just to dictate this
5:25rather than type everything out. It lets
5:27you talk all about your business, all
5:30about your clients, and then you're not
5:31going to censor yourself the way that
5:32you would would if you were actually
5:34typing it out. So, I would just click
5:36here and I would say something like, uh,
5:39I need your help in coming up with the
5:42first 10 YouTube videos for a tattoo
5:45artist. Um, he wants to talk about
5:48things like, you know, how to prep
5:50people for their first tattoos, things
5:53they need to know. Use the Vid IQ
5:55connector. And I want you to really
5:57figure out um what's working in the
6:00tattoo space right now on YouTube. What
6:02has search volume that we could go for
6:04for especially for a new channel where
6:06they would actually be able to rank for
6:08these terms and then balance that out
6:10with titles that actually are a little
6:13spicy that has have something to say
6:15beyond just the search term.
6:18Click there and we can see the Vid IQ
6:21connector. The little logo is right
6:22there. So then at this point I would
6:24just let that run and we're going to see
6:28what it comes back with. Now is a good
6:29time to talk about usage credits. So
6:31you're basically going to be going
6:33through credits in two places. One on
6:35ChatGBT itself and two with Vid IQ. It's
6:39all kind of a creditsbased system. So
6:41you pay for ChatGBT, you get a certain
6:44number of credits to use a week. And I
6:46won't lie, Astra does use them pretty
6:49quickly. So, this isn't the kind of
6:50thing you would want to do every day,
6:52but it's the kind of thing you could do
6:54once a week or you could do once a
6:56quarter just to give yourself like a
6:58bunch of topics. You can ask for as many
7:00as you want. And then when we get to the
7:02Vid IQ side, it depends what plan you
7:04have there, too. Um, I did go through a
7:07lot of credits yesterday when I was
7:08testing this. So, you just want to be
7:10mindful of that. But, if it's like a
7:12once a month thing that you're doing,
7:14it's not the end of the world. It's a
7:15good use of your credits. So, from here
7:18on out, we can just see what it's doing.
7:20It's pulling up all of these uh searches
7:22within Vid IQ. Um, I'm just going to let
7:25that run and then we'll see what it
7:26comes back with. Okay, so it worked for
7:2817 minutes and 21 seconds. So, and
7:31that's a big brain working for that
7:33long. So, we'll see what it comes comes
7:34back with. We can see all these like
7:37searches it's done with Vid IQ. We'll
7:39just get to the good stuff though in the
7:40end. Um, let's see what we've got. So
7:44says, "I'll I build this channel around
7:46helping people spot problems before they
7:48become permanent tattoos." Cool. So it's
7:51a lot of background it's giving you,
7:53too. So it's telling you what's on its
7:56mind, why it's kind of going where it's
7:58going. So you would want to read through
8:00all of that. And it says these are the
8:0210 I'd launch with roughly in this
8:04order. Um, all right. Cool. So let's
8:07let's go. We got tattoo placement
8:09mistakes you only notice too late. Um,
8:12and the search demand is the word tattoo
8:14placement. It gets about 4,000 searches
8:16a month. And then it gives a little more
8:18detail here. Then how fine line tattoos
8:21age, the photos you need to see. Cool.
8:24It's got a search demand behind that.
8:26And it's also got evidence of other
8:28videos that have done well. So, it's
8:32kind of basing it on that, too, which is
8:34really smart. We've got another eight
8:36tattoo styles, one idea. Which should
8:38you choose? How much do tattoos cost?
8:40Price. I price the same idea three ways.
8:43So already I would say these are pretty
8:46good. It's pretty solid choices with
8:48some good solid thinking behind each
8:50one. But now let's go to our next case
8:52study that was submitted. So this is
8:53Christine Harrington. She's a sales
8:55coach. And if we look at some of her
8:57videos, we can see her views are not
9:00that great. And for something like this,
9:03like a sales coach isn't that niched
9:05down. Like there should be a little bit
9:07more views on on videos like this. So,
9:10we have 26 views, 78 views, 76 views.
9:13Um, and she's been at it for 12 years.
9:15So, let's see if we can get her out of
9:18whatever rut she's in. Um, with some
9:21some more thoughtout topics because if
9:23we look at these, we have the mental
9:25game of sales. How to win every time.
9:28Forcing a buyer to buy is not a sales
9:30strategy. Here's what will happen. And
9:31so I feel like some of these are just a
9:34little too kind of abstract or this one
9:37might be a little too specific and it
9:39might feel like it gives the whole thing
9:40away. So let's go ahead and we'll go
9:42back to chatbt and again I'm just going
9:44to dictate. I'm just going to click on
9:46the microphone button. Help me make the
9:50best list of 10 YouTube videos for my
9:52friend Christine. She's a sales coach.
9:55She mostly deals with people who are
9:58soloreneurs who haven't uh gotten into
10:01sales yet. They've never made sales for
10:02themselves or they are sales
10:05professionals themselves who are kind of
10:07struggling with that. I'm going to paste
10:09in a link to her YouTube uh profile so
10:12you can look at her videos and you can
10:13see what's done well for her, what
10:15hasn't done well for her, and take a
10:17look at her competition and see what's
10:19working for them. see whatever uh
10:21keywords that have some good search
10:24volume that she could use. But
10:26basically, she's been struggling at this
10:27for a while. So, we want to make sure we
10:29get her a win and her next 10 videos are
10:31going to perform higher than her
10:33average. And then I will just paste in
10:36her YouTube link. And again, we want to
10:39make sure Astroax is selected. And we're
10:42going to click on the button to send it
10:44off. Okay, so it's done. It worked for
10:4615 minutes 44 seconds. Um, we can see
10:49all these searches again. So, let's get
10:52to the good stuff, what it's
10:54recommending.
10:56So, I would build these next 10 around
10:58specific moments where she loses a sale
11:00and show her handling those moments.
11:02Okay. Uh, her channel already supports
11:05this direction. Cool. So, it's giving
11:06all the cool background. And then here
11:09are the 10 titles. Uh, what to say when
11:11a prospect says I'm not interested. Live
11:13roleplay. So, substantive sequel. So,
11:17basically her uh one video, how what to
11:20say when someone says, "I'm not
11:21interested," got way more views than her
11:25normal videos did. It got 35 times her
11:29normal views. So, it picked up on that
11:31and said, "Maybe we make a sequel to
11:33that." So, that's awesome. Stop being
11:35ghosted after sales calls. And it's um
11:38based on this guy Ryan Dorne's Stop
11:41being ghosted after a client sales call,
11:43which was an outlier for him as well.
11:46So, these are pretty well thought out
11:48and they're based on evidence again. So,
11:51um I like all of these, but I have to
11:54wonder, we gave it so few guard rails
11:57and so few things to look for and
11:59personally I know what goes into making
12:02a good YouTube topic. Like, there's a
12:04bunch of different factors. So what I
12:06did is I made an actual chatbt project
12:10with all of my rules for the road
12:12basically of what kinds of videos you
12:14need at what stage in your YouTube
12:16journey. So I packaged up that project
12:18and all of the instructions. It's called
12:20the expert topic research kit. So get
12:22that below and all the instructions are
12:24in there. We're going to use that now to
12:27see with all of that extra context, can
12:31we get even better topics and titles
12:33picked out? And by better, I just mean,
12:35are they going to have a better chance
12:37of working at whatever stage these
12:39channels are in. So, um, let's go ahead
12:42and go back to Chat GBT and we'll go
12:44over to projects. So, you'll install
12:47this on your own. All the instructions
12:48are there. It's really easy. And I'm
12:50just going to pick, uh, YouTube topics
12:52test. That's the one I've been using.
12:54And now, I'm going to prompt it the same
12:56way. I'm just going to dictate it. I
12:57would do it the same way I would tell it
12:59about the person. The only difference
13:00now is it knows all the rules to come up
13:02with like the best possible topic. So I
13:04just do that again right here for
13:05Christine.
13:08All right. So here's what it came back
13:09with this time. It says I'd build three
13:12videos around a promising format on
13:14Christine's channel. A familiar prospect
13:16statement followed by exactly what to
13:18say back. So this is going on her, you
13:21know, what to say when someone says I'm
13:23not interested. Remember that was her
13:24her big hit on her channel. The other
13:27version said like make a sequel to that.
13:30Essentially, this is even smarter
13:33because what it's saying to do is it's
13:35basically saying make different versions
13:38of that. So, what to say when someone
13:41says I need to think about it. What to
13:43say when someone says it's too
13:45expensive. What to say when someone says
13:48just send me some information. So, these
13:50are like really common sales call
13:52objections and she's just going to apply
13:54what worked to all of those, which super
13:57smart. And again, we've got all the
13:59context of like why it's suggesting
14:01these. So, then we have how to write
14:03cold emails without begging for 15
14:05minutes, 40 years in sales, how I'd
14:08build a pipeline from scratch, stop
14:10getting ghosted, fix your sales
14:12follow-up process. So yeah, I would say
14:14this is a huge step up from what it was
14:17just giving us originally, but that's
14:19for a case where there's plenty of
14:21things to talk about, right? There's a
14:22lot of sales people out there. It's easy
14:24to make a lot of different videos for
14:26it. What if you were in a niche that is
14:29just way way smaller and you're really
14:32struggling for topics that anyone would
14:34be searching for and anyone would, you
14:37know, have an appetite for. So now I
14:39want to talk about one of my students,
14:41Barbara. Uh she sells something way more
14:44specific and niched down. Uh basically
14:47she consults with people how to come up
14:50with the best possible branded merch for
14:52their business to as client gifts to get
14:54clients to keep clients that kind of
14:56stuff. Um and her topics have been a
14:59little scattered. So how to turn your
15:00brand into a product people actually
15:02want the junk drawer test every merchant
15:05buyer needs.
15:07So, these probably aren't getting a lot
15:09of search volume um as judged by her
15:12view counts at least. So, I think this
15:13might be a huge help for Barbara to
15:15finally find the right topics. So, I
15:17just used the same project within chatbt
15:19and had a conversation with it about
15:21Barbara, what she does, who her clients
15:23are, and this is what I got back. So, um
15:26the first one, seven corporate gift
15:28ideas that don't look like corporate
15:30gifts. I think that's great. And then um
15:33the phrase corporate gift ideas has
15:36verified search demand. So that's
15:38awesome. So again, this is based on
15:41evidence. It's based on research and
15:43data, not just feelings. Um the bulk
15:46discount that makes client gifts cost
15:48more. Hm. That's interesting, right? You
15:50wouldn't think a discount would make
15:52something cost more. So it's got some
15:53like drama behind it, which is good. And
15:56it's also, by the way, saying what you
15:59could make the video about. So, it's not
16:01just like saying a topic in a title.
16:03It's giving really good advice like what
16:06exactly might go in a video like this.
16:10Then we've got company swag I'd pay for
16:12and what I'd leave behind. I love that.
16:15And then yeah, company swag has verified
16:18search demand and represents a real
16:19purchase Barbara can help someone make.
16:21So, these are really solid topics that
16:23it would have been really hard to come
16:25up with before we got this upgrade in
16:27Chad GBT. So, and if you want to get
16:29that expert research kit, it's for free
16:31right here. So, just click there and uh
16:33I'll send it right over to you. It's
16:35really the best way to make this work
16:39for you in literally less than 20
16:41minutes. So, get that and I'll see you
16:43in our next