Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
She's making $10 million per year in revenue and she doesn't
exist. What?
Yeah, they scaled from zero to 20 million in annual recurring
revenue just with AI ads. The app didn't exist 12 months
ago. If.
You were to be a user trying to make 10 grand a month as soon as
possible and you didn't have much money to work with.
Which path would you take? If you want to do 10K like I
mean that's easy, like really easy.
(00:21):
I will create 20 ads a month foryou for $2000.
These people will have the budget because they already
spent 500K per month on meta ads.
I will just do that for 100 companies.
That's probably the easiest way to get to 20 KA month.
What is this world coming to? Did you guys know that there are
(00:43):
AI influencers making millions of dollars a year?
Yes, that's real. Some dude in his mom's basement
is generating AI images and videos and posting them to
Instagram, Tiktok and everywhereand getting big huge sponsorship
deals. So with this new AI short form
video generation technology blowing up like crazy, how do
you profit from it? Because we can shake our heads
(01:04):
at these AI influencers or we can go create some and monetize
them in our paid ads, in organic, social, ETC.
So I had Romaine Torres on todayand he is one of the biggest
experts in the world when it comes to AI generated video
specifically for making AI ads and AI influencers.
Your brain might explode on thisone.
Please enjoy. Initially the main use case was
(01:26):
paid ads, but very quickly we realized that was a huge
opportunity to create these AI influencers that will basically
live on Instagram. Maybe I can share my screen and
show a couple of examples, please.
But you know, there is this famous influencer right here,
Michaela. Apparently she's making $10
(01:46):
million per year in revenue and she doesn't exist and are more
all of that. Yeah, she's girl.
All the pictures that we see here all are AI generated and
Bronze are paying hundreds of thousands to get a story from
her or just a real. And really the limitation
(02:11):
recently was you could build very good static images of AI
inferences, but having them talkwas very hard.
And that's basically what we solved.
And a lot of people now have used our technology to create
either a clone of themselves. So let's say you Chris, you want
(02:32):
to create content on Instagram and you want to grow your
Instagram channel, but you don'twant to spend too much time
talking to the camera. You could create clone of
yourself with our cads and make video reels where you will do a
video edit. And at the bottom of the video,
for example, you'll see your clone and at the top of the
(02:53):
video you'll see a video about whatever you are talking about.
And as you can see, these videosdo pretty well, right?
2 million views here, 2 million views.
And all of these videos were created using his AI avatar
concept to working prototype in just seven months humanoid.
And obviously there is a lot to be done, like to make the video
(03:14):
very dynamic and stuff. But yeah, you'll basically save
the time of talking to the camera using this technique.
How much of your tool helps generate the video and the voice
and how much of it like helps actually edit the video?
So all kinds of reports that simple.
Creating very good and very realistic talking actors.
Maybe I can just do a demo of the tool by the way, and just
(03:37):
show you how it looks like very quickly.
And so so it's clear for the audience.
We really focus on like helping you writing a script, selecting
an actor. And this actor can be either
yourself or another AI actor that is from our existing
library and generating the video.
So basically the way this will work is you will type your
(03:59):
script right here. You can add tags to control the
emotion because what we realizedin our data is that the voice
has a ton of impact on how real the AI avatar will look.
If the voice gets too robotic, it's going to be very hard for
people to engage with the content.
So you can add tags with the emotions you want.
(04:21):
You can even control the accents.
You can like add A tag that says, I don't know, French.
I don't know why I would want that, but maybe you want French
accent and then the AI voice will basically talk with a
French accent. So you will like write scripts
and then select an AI actors through our library of thousands
(04:43):
of existing AI actors. Or you can just create your own.
So you can just click here type a description of the perfect AI
actors. So I don't know a 23 years old
female inference or in a coffee shop, you can just iterate and
start a first version a person based on the prompt.
(05:06):
And then you will be able to edit this person with perfect
consistency. So you are going to be able to
generate her in different situations, different
backgrounds. You're going to be able to
change the color of the hair, change every single detail that
you would want to change to get the perfect AI influencer.
(05:27):
And then when you are happy withthe image that you generate,
because it will start with an image, you can then turn this
person into a talking actor. So you would write the
description of the AI actor or choose an existing AI actor in
the platform, write a script, listen to the audio, and then
basically click generate and youwill get a video of someone who
(05:47):
will be talking to the camera with like perfect lip sync and
motions. So my connection is really bad.
Yeah. Oh wow oh wait, am I really AI
generated? And then basically you can like
have the person in tons of different situations do all the
scripts. That's right.
And then Nicola here, create videos for social media.
(06:10):
And at some point, when you get a big enough audience, brands
will come and pay you to create content to promote that stuff.
Now you and I see these videos and we can tell it's AI.
What? What percentage of the general
public do you think can tell? And if so, does it really matter
that much? So it's a very good question.
My initial answer was the younger the audience, the the
(06:34):
more that will notice it's AI. Right now we have a lot of
clients and that's actually a big secret, a lot of clients who
target older audiences with withour content because if you are
55 years old, there is a much higher chance that that much
interested in technology and AI and you will engage much more
with this type of content. So that was my, but really the
(06:57):
technology is progressing super fast.
And there is this tech influencer right here, Olivia
Moore. She's like a partner in the top
USVC fund. And she actually yesterday or
two days ago, she made an experiment where she created an
account on Tiktok on a very specific niche that will
actually target very young people.
(07:17):
And she created an account for the niche.
It's funny, she's quote at the University of Alabama because I
was there yesterday on campus. I saw people who were following
this AI profiles that she created.
So basically she she created a aprofile about like admissions at
(07:38):
Alabama's. She started posting content on
TikTok with this AI influencer and 20 years old girls were
commenting stuff like, hey, you will get there.
No worries, you will get your admission because she was making
a video where the AI and we can see some examples here.
And actually the account is public.
(07:58):
We can find her on TikTok and she makes a lot of views.
Some of her videos make like hundreds of thousands of views.
And she did this experiment veryrecently.
And it's very interesting the way she created this.
By the way, she she uses this tool from TikTok where you can
see for each keyword how many posts are done and all the
(08:20):
hashtags trending. You can basically find any
hashtag that you aren't interested about and see there
is an opportunity to create content and just start creating
content for this niche. And in a matter of weeks, she
managed to get a ton of views. But yeah, it's kind of insane
how how much you can. That's insane.
(08:41):
I. Think it's going to just get
better every every week. If you look at AI six months
ago, like honestly the progress has been insane.
So I think that's a huge opportunity, easier to create
your own AI inferencer in a veryspecific niche like she did here
(09:01):
or clone yourself if you are already a content maker and you
want to post more content. And maybe the third big
opportunity here will be go findexisting inferences who have
accounts. People like you basically say
and say, hey, Chris, you want tocreate more content.
Everyone who creates contents wants to be creating more
(09:22):
content. How about instead of you
recording yourself, I will create content for you with your
avatar and you will validate thethe script.
You will validate everything. But instead of having to record
yourself, I'll just like create the content with your avatar for
you. Do you have customers that are
using it for that purpose for asan agency basically?
(09:43):
Yeah, a ton, a ton of customers.What are they charging?
Have two types of cases. The first type will be seeing
like doing this service for influencers and typically they
will charge less because influencers will have a lower
budget. But the other big use case is
findings a company that is already spending money on ads
(10:05):
and helping them create content.So I'm going to show you what I
do here with an example SO. Like look kind of scrolling on
the Facebook ads library, seeingpeople have a lot of active ads,
reaching out to them and saying hey I can help. 10 extra output
with AI generated ads more creative.
Exactly. Yeah, that's about, that was
exactly what I was about to to show.
So on the Facebook ad library, which is the public data of like
(10:27):
all the ads active right now, you can find any app, any kind
of product depending on what you're interested about.
So obviously I'm interested in tech, so I'm going to be 10 like
bias towards apps, but I'm goingto also show you an example of
the more traditional business where you can do that.
So find an app, this app, for example, they scale from zero to
(10:49):
I think like 12 or 20 million inannual recurring revenue is the
type. It's public.
We can look at it just with AI ads.
Basically, they created the app 12 months ago.
The the app didn't exist 12 months ago.
They only started the app run it, they started running
Facebook ads with AI inference just created with Arcads and
(11:11):
they scaled more than 10 millionin revenue only through data.
And as you can see they are testing a ton of ads.
So they are like testing 650 adsright now.
This is really interesting to me.
I think that it's critical that when people do this, they pick
the right niche or the right industry in which AI generated
ads are more likely to work better, right?
This is an English learning app,which means they're probably
(11:33):
showing this ads to an international audience.
I'm guessing that the international audience is less
AI savvy. They're not as AI sensitive like
AI generated video sensitive, therefore the ads are more
effective. Or maybe they think the ads are
automatically transcribed into English or into whatever
language and they're used to seeing that because English is
not their first language. Would you agree that like using
(11:56):
AI generated ads in the right industry could have like a
multiplier effect on how effective they are?
That's a very, very good point. Another example of that will be
this company that are running a ton of AI generated ads.
So it's called disability path and basically they serve.
(12:16):
So maybe I can show you an ad. By the way, this is an ad that
they created with archives. $4018 a month to cover your
grocery bills. Yes, you heard that right.
If you're over 50 and can't workbecause of a health issue, you
could be eligible for up to 50. So yeah, these people they
basically sell leads to these disability companies where they
(12:40):
do the admin work for you to getyou these low ones that you can
get if you have an A disability and you are over 50 and these
people needs to get leads. So basically need people over 50
to sign up to their services andpeople over 50, well, most of
the in most cases just like not be triggered by the fact that an
(13:02):
ad is AI generated. Honestly, even in some ads, you
or I, we cannot even see that it's AI anymore.
Like we are getting to this point already.
It's it, it requires some skills.
And we can even talk about that if you want to at some point.
What are the skills that helps you make a very good AI ad that
make it very look very realistic?
(13:23):
But I'm breathing like in six months, like everyone will be
able to do a perfect like 100% perfect AIUGC ad where it's
going to be impossible to distinguish between a real ads.
But right now I agree like 100%.It's like the biggest
opportunity is to go find these types of companies.
Yeah, I look at this in like 2 different ways.
(13:45):
So let's take a super competitive category to a
somewhat sophisticated audience.Creatine gummies, right?
There's hundreds of if not thousands of Shopify stores
selling different types of creatine gummies that are
probably all made in one to three different factories,
right? If IA 38 year old white male in
North America see an AI generated video ad for creatine
(14:06):
gummies, I'm likely to be uninterested turned off, you
know, because it's like low effort AI generated.
Nope. Whereas if I see like a more
professional ad that's, you know, handmade for creatine
gummies, I might have a bias to click on it more right?
But let's say I'm looking for and this is something I actually
ordered from the Internet. I ordered trees like I literally
(14:28):
like 4 foot tall trees I orderedfrom the Internet from South
Carolina and I saw Facebook ads for them.
If I was that tree company and AI generated video ads helped me
scale my creative 10X. If I really need a tree from the
Internet and I'm having a hard time finding it and they're able
to make 10 times more ads because of AI and that puts it
(14:48):
in front of me or it's more likely to get in front of me,
then I'm going to buy it, whether it's AI or not or not.
Just like I listen to what ChatGPT says because it answers
a question I need, even though Iknow it's ChatGPT, right?
So there's kind of two schools of thought, like, oh, they're
not going to take it seriously because it's AI.
And then the second school of thought is doesn't matter if
it's AI, if you're solving a problem, they're going to click
(15:10):
and or they're going to buy, right?
Yeah, I think that's the whole story of advertising.
Like when you watch a video ad from Charles Clooney telling you
Nespresso is the best coffee in the world, and do you really
believe Charles Clooney will drink Nespresso every single
morning? No, but it doesn't matter
because now you know about Nespresso.
(15:30):
Now you saw an ad that is entertaining, that is a good
story about Nespresso that is funny.
I think to your point, I really agree that people who sell
services where it doesn't reallymatter if you trust or not, the
person who will promote the service will have better results
for sure. If you are selling, buying
create creating game is I I guess you would want to have
(15:51):
someone who has some sort of knowledge about the topic
telling you to buy them, right? You want to have a doctor?
You want to have maybe an influencer you love talking
about it. If I saw like Matthew
McConaughey ad for creatine gummies, it's like, yeah, he's
fine. I could take him or leave him.
Not my favorite actor, not, you know, not the worst actor.
But if I saw Tim Robinson, like my favorite comedian actor, do
(16:13):
an ad for creatine gummies, I use these.
I like these, I'm going to be more likely to buy just because
I like him. It's the same product, right?
And so the same principle applies for AI generated versus
otherwise. Like in this example, Tim
Robinson is higher quality for me because it's more relevant.
It's there. Those are synonyms in this
context, right? Just like AI generated is lower
quality AKA less relevant. And I think also something
(16:36):
that's going to happen to our initial point on AI influencers
is at some point some AI influencers will be so good at
creating content that people will actually be attached to to
them, like people are attached to Spider Man, right?
And he doesn't exist. But it's a good point.
Yeah, I. Mean you can be very attached to
(16:58):
to personalities that don't exist.
I sorry, I think it's interesting, the first two
examples you showed me of like these AI generated ads that are
crushing it. Are those two international, you
know, English as a second language speakers and people
with disabilities, right? Like I think there's a
correlation there that both of those categories of people would
(17:19):
be less likely to recognize it by no fault of their own.
It's just not something they come across on a regular basis.
Whereas 30 something year old white dude playing around with
ChatGPT and Grok all day. It's second nature for us to see
an AI generated video, you know.That's a very good one.
And there is AI think 1/3 category which is very
interesting. And unfortunately I I won't be
(17:39):
able to show you this example because the person asked me not
to because they are making so much money, OK?
OK. Which I guess is a good teaser,
but so it's products where people are already aware that
the service I will buy is AI. And what I mean by that, and the
example I had in mind here is you probably saw these new AI
(18:04):
voice agent technology companies.
So basically you could go, and Ithink it's a very good business
to start. When I see the numbers of people
who use archives to promote these, you could go say, OK, I'm
going to create a voice agent that will do appointments for
hairdressers, for example. That's probably an idea you
(18:25):
already discussed on the becauseI don't know.
But if you are hairdresser, people call you to get
appointments or if you are dentists, people call you to get
appointments and you have to paya secretary to answer those
calls to book and everything. This can be done by an AI voice
agent. Now, there are a lot of
technology and software that lets you do that.
(18:46):
So a lot of people have packageda service on top of these AI
voice agents companies. They package the service and
they found dentists or hairdressers or whatever the
niche they are targeting and they sold a service, a package
service where they will tell them you don't have to do
anything. I will just replace your
secretary for like 1/2. And like not even half like a
(19:08):
tenths of the cost. The only difference between the
person who will be taking the appointments will be AI, right?
And so we have a lot of customers who have created a
service like that and then they run paid ads to acquire
customers. So let's say you are sitting in
this. I've recently heard this called
dog fooding. You're dog fooding it.
(19:28):
You're eating your own dog food,right?
You're using your own thing to sell your own thing.
Therefore all cards are on the table.
No one's turned off by it because your ad is showing that
it works. Exactly.
That's the point. Yeah, I think that's the truth.
And then by the way, that the AIlanguage running up.
So it has both actually, becauseit targets people from other
countries who don't speak the language perfectly, but also it
(19:51):
uses AI to train you and learn this new language.
So yeah, I think that's a very good point.
Like choosing the right target audience will increase some of
the likelihood of the ads doing well, basically.
What other kind of trends are you seeing from the back end of
your company that people could could learn from like industries
(20:12):
or offers or the duration of videos that people find the most
success with etcetera? Yeah, that's a very good point.
So and by the way, we mentioned like the possibility of creating
your own service, like just creating a service and selling
it's through the AI ads, but youcan also just sell the service
to other companies. That's probably the easiest way
(20:35):
to get to, I don't know, 20 KA month because you were asking me
how much people would pay for these AIUG figures.
You could reach out to learn howthis language learning app and
tell them, hey guys, I will create your ads, the ads for you
and I will like package it like,I don't know, maybe that's 20
ads. I will create 20 ads a month for
(20:56):
you for two, $2000. So you all pay me basically $100
per ads and that's already goingto be cheaper than working with
actual UGC creators. These people will have the
budget because these people, they already spent 500K per
months on Meta ads. So for them spending 2K is like
less than 1% of the ad budget that they have.
(21:19):
So that's a big, big trend that we are seeing.
People are subscribing to our software to create these ads,
developing the the skills that it takes to make very good ads.
And we are going to talk about that like the trends we are
seeing right now, developing theskills to make very good ads and
then sell that as a service to other companies who already have
(21:41):
a product. You don't have to even build
your own product. You can just like serve existing
companies who who have needs. How can you see that they're
spending 500 KA month? So you don't have public data
about the spend of an app, but you can see them revenue.
And it's very, for example, in the app industry or mobile apps,
(22:01):
it's a pretty common knowledge that people are spending 70
percent, 60 to 70% of their monthly revenue in paid ads.
Because when you create an app, you have very low costs of
running the app. The main cost is, but if you
look at this language learning app we were just talking about.
So if you see, if you go to thiswebsite, Sensor Tower, it's
(22:23):
probably data, you'll see the revenue that they do on a
monthly basis on iOS and Android.
Very interesting. I've never thought to to do
that, to take their monthly revenue and just multiply it by
.7. That's very smart.
It's very specific to the app industry.
I guess it can also be true for Shopify, maybe.
But yeah, apps are public. The revenue is public.
(22:44):
So that's very that makes it easier.
And there are a ton of apps you can reach out to and sell these
services too. It's becoming more and more of a
a real skill to be able to use AI the right way to create ads.
And you were asking me about thetrends, the techniques to make
very good ads that will perform.Really, AI can pretty much do
(23:04):
everything you imagine now if you use it the right way.
So I don't know, an example of like something that will make a
lot of difference will be how doyou make the first second of the
video visually appealing? And how do you make sure that
when people are scrolling, therewill be interesting to watch
more of the video. And one way to do that is to
(23:27):
have an AI actor dressed in a funny way.
So we took this AI actor that isalready in the platform right
here. So that's a traditional video we
will do with this AI actor. Hey, you're looking at your
camera. Did you know I'm not real?
Now that one looks very real. That sounds real too.
(23:49):
Yeah, he's pretty good, right? That one's legit, yeah.
So a couple of things you can do.
You can do a gesture, a movement, and we can we can do
that with the platform. So you can say at the beginning
she points at the camera becausewhen you say it looks real, it's
obviously because of the looks, but it's also because of the
gestures and the voice. So you can say the beginning,
she will be pointing at the camera.
(24:10):
Hey, you looking at you. And that's what we call a visual
hook. So you will increase the chances
much of the person like stop scrolling because you kind of do
this gesture. Another way you can do that is
you will take this actor and then now you will say he's
wearing a cosmonaut Nada Benada just scope released by Diego.
(24:31):
And I mean, probably this cosmonaut ID is not the best ID,
but at least it's something thatis very unique that will make
people have an additional chanceof stop scrolling and then AD
test the same thing where instead of like wearing a
cosmonaut suit, he will be wearing like kind of and Indiana
(24:52):
Jones style suits. Have you seen that guy on
Instagram that has his videos? He's like laying down on a
skateboard rolling down a hill. Have you seen that?
No, actually, no. The camera is positioned on his
chest and he's laying on some sort of rolly device like a
skateboard and going downhill ona sidewalk and like screaming
(25:14):
into the camera about something irrelevant to what he's doing.
And his videos get millions and millions of views because the
visual hook A and the retention B because you just, you're
thinking he's going to crash thewhole time, right?
And I think people making content underestimate how
important their setting is, right?
Like just a snake around your neck or just something for
(25:35):
someone to look at as you're talking is so critical.
That was one of our clients who did an ad where the AI actor was
driving and they were not looking at the road.
Everyone was like, when are theygoing to start looking at the
roads? And people got aware that they
(25:56):
got an accident. So you can do that with AI.
Obviously you don't want to be recording that in real life
because it's so dangerous, but you can do that with AI.
And that's one of the things with AI also, you can just do
things that would require an entire production team.
Not only can you do like just test different audiences, if you
are a 25 years old and you want to target people who have
(26:19):
disabilities and are over 50, obviously you are not going to
be very relevant when you recallyourself and do a video.
But now you can use an AI actor with a filter and say I want to
have this guy who will be, I don't know, senior.
And you will only see AI actors who are over 50 years old or 40
(26:40):
years old. And interesting.
But also you can just like do something crazy if you don't
want to target a different audience and yourself, but you
just want to catch the people attention.
Maybe you can create an avatar that is on the skateboard as you
said. And as they are talking, they
will be riding on the skateboardand any just be creative and
(27:00):
then it's about like doing this unique thing that the other ones
don't do. If you were to take your
customer base and put them into like 3 to 5 different buckets of
like, all right, these guys use them for AI influencers, these
guys use them for AI generated video ads.
What would those buckets be? What are people using your
service for? The first one is mainly ads, so
(27:21):
promoting either their own services or selling AIUDC to
other people. The second one is AI influencers
and inside if we take ads insidethese buckets, we have three
main verticals really the first one will be mobile apps.
We have a lot of them. People would create an app on
(27:43):
novable and then guess what, youneed to get users to the app.
So you need to run ads and to ina very competitive world, you
need to be very creative about the ads you do.
So the first one is mobile apps.The second one is E com.
So you are selling a physical product and by the way, you can
do that with AI. You can have them like actually
(28:05):
hold your product while they talk about it.
So the second one is E com and the third one is services like
the disability path I showed you.
It can be lead generation in a very broad sense where you will
sell leads to any type of services.
So hair salon, dentists even entrances people will like run
(28:28):
ads on like why you should like try this new entrance and then
sell the leads to insurance companies.
OK, when it comes to AI influencers, I have to think
there's a lot of survivorship bias where it's easy to point to
the handful that have millions of followers making all this
money but nobody sees or knows about the ones that just never
(28:49):
did well, never got any followers.
It didn't work. Where are people failing when
trying to create and grow AI influencer accounts and what
could they do better? So I would say lack of
consistency and lack of creativity.
It's very similar to content creation without AI.
Right now, probably thousands ofpeople who would want to be
(29:13):
Youtubers or Tik Tokers and create content, but they only
post a couple of weeks and then they give up.
That's the main one. With AI, the same principles
will apply. If you just go ahead, try it for
like hours, generate one or two video, post them, I can
guarantee that there is 99% chance that they will not
(29:34):
perform. And then the second thing is
creativity, right? When you create contents, you
need to be creative. You need to have good ideas.
I showed you the example of the BC who created content for
sophomores in the University of Alabama.
She was very smart about pickingthe right niche.
So she saw that there was this trend right now on Tiktok
(29:56):
because Tiktok lets you find this with the TikTok creator
search inside, which is an inside like a, a feature inside
TikTok where you can see the trends that are rising right
now. So she was very smart in
choosing the right trend. And then she was very smart in
the content that she did. She was very creative like by
having emotion in the script. The top video was a video where
(30:17):
the AI in transfer was crying. She also played with the looks
of the person. And maybe one last thing is,
which actually will be contractual contradictory to
what I just said, but I think a lot of your content will be
looking at what's already doing well and trying to replicate
that with AI. And that's one of the key
advantages of AI, right? You can like take something
(30:39):
existing and redo it in a different way.
I think the creativity part is is super key.
Like people from what I've experienced, people that want to
create content, they don't want to show their face.
No one wants to, right? Like that's normal.
And so like, Oh my gosh, AI influencer, that's exactly what
I've always been looking for, right?
Because I don't have to show my face.
It can be a different person or it can be an AI generated
(31:01):
version of my face. And so they think that now they
found the solution to all their problems when really they need
creativity. They still need a good hook,
they still need something to keep people retained throughout
the whole video. I think people also get hung up
on consistency. They're like, if I'm just
consistent, if I just put out one video a day or two videos a
day, I'll get there. It's just a matter of time.
(31:23):
But if you're not improving every video or if you're not
being creative enough for if youdon't have good hooks, if you're
not learning from your mistakes,there's a lot of accounts out
there with thousands of videos and 82 followers because you
know, consistency is just one factor.
You've really got to become a student of the game, like you
said, learn what's working with other people and be super
creative whether using AI or not.
(31:45):
And to be tactical, the tools you can use to that.
So if you are doing ads, basically Facebook ads every
just looking at the ads the other people in your industry
are doing, you also have some tools built on top of the
Facebook ads every like foreplayor you can create boards and you
can save the ads that you see from your competitors.
Basically create inspiration. The Tiktok search creator
(32:08):
insight is really, really good. It will literally tell you which
pieces of content are lacking from the platform.
It will tell you, OK, so many searches on this keyword and
only this number of videos. So if you make a video about
this very specific topic, there is a very high chance that it's
going to do well. Also going on Tiktok, Instagram,
(32:31):
if this is what you are doing, downloading a lot of videos of
people who did well, you know, there is this concept of
outliers. So basically you will take a
TikTok account, you will look atall the videos that they posted
and one of those videos will probably have done 10X better
than the other ones. Like why did they do 10X better?
And then bring them to a Google Sheet and try to replicate them
(32:53):
with AII truly believe the revolution of AI is you can
create more faster, you can testmore things.
And also it opens up the opportunity for people who are
not necessarily the most good looking, who don't necessarily
speak the language perfectly. And we'll see more and more of
these accounts. Knowing everything you know now
(33:14):
about your own software, how people are using it.
If you were to be a user of yourown product and login, create an
account and trying to make 10 grand a month as soon as
possible, what would be and you didn't have much money to work
with? What would be your go to action
items to use your own software to get to about 10 grand a
(33:34):
month? Which path would you take?
Yeah, 10 grand. And let's say I'm not like tech
savvy and I cannot code. I cannot even like use my
colleague and I don't want to doa tech product.
You got 1000 bucks to work with?Yeah, I mean, I will just like
go to the Facebook Ads library. I will open the Facebook ads
library, look at companies that seem to be running lots of ads,
(33:59):
find their LinkedIn accounts andI will target like smaller
company because you don't want to be targeting companies with
200 people plus with a lot of can't.
Find the decision maker. Exactly.
I would want to to basically be able to reach out to the CEO.
So I will find out LinkedIn. I will then go to our CAD,
create an ad very similar to onead they are already running.
(34:23):
Maybe they are running existing UGC ads with UGC actors.
I will tell them OK, this UDC adthat you probably paid $500.00
for and that took you 3 weeks toto get, I will do it for you.
And actually the first one will be free here.
Is it here. It is like I, I did this video
for you and I, I did it with AI.That's the only difference.
(34:45):
But I did it for you and I will just do that for 100 companies.
And then if I do that for 100 companies, I will probably get
calls, like phone calls with 2030 of them and then a third of
them will pay for my services and I would sell them the
services. 3 to 5K per month. Yeah, that's already 15K in
(35:09):
revenue. And I mean, I think in I can
pull this off in one month. In one month I can be making 15K
in revenue doing that with like 0 network just LinkedIn and
reaching out to people. That's perfect.
Yeah. It's just a it's just a numbers
game. You need more shots on goal.
If you want to really make $100,000 in revenue, probably
(35:30):
build the product or create an AI influencer and try to become
the next Nikolao because I'm going to be a ton of them in the
next years. If you want to get 3K or a
million a month, if you want to do 10K, like I mean, that's
easy. Like really easy.
I love it. You could do both at the same
time. Create an AI influencer and
start, you know, start an agencyat the same time.
(35:52):
Why not 100%? Yeah.
Is there anything I didn't ask you that I should have asked
you? Like anything interesting about
how this works or any interesting?
I think this has already been amazing, but is there anything I
missed? We can talk about like the
advanced level. I mean, I don't know if this is
something you guys might be interested, but the like level
(36:13):
one will be you go to the platform, you create ads and you
write your scripts and you brainstorm, you create the
perfect actor level. Advanced level would be OK now
actually, instead of like me brainstorming IDs and me
creating one or five videos a day, I will create Google Sheets
(36:35):
with lots of existing ads. Send this data to ChatGPT.
Let ChatGPT read the transcript of the transcript of these ads.
Ask ChatGPT to recreate a new script for another for my
service that I'm setting and then connect to our API through
an automation tool like zapyourmake.com NA 10 gum loop
(36:58):
and create these ads automatically.
And instead of like making one to five a day, I might be doing
100 to 500 a day. And then I would like really
increase the chances of finding this viral ad that will bring a
lot of profit, not by brainstorming, but by letting AI
do the whole flow. This is a little bit more
(37:20):
advanced, but I'm seeing more and more people getting insane
results with with with these flows.
And that's probably a very big opportunity as well.
Yeah, because everything you described with that automation
example could could be automated, right, Or not with
the automation, with the agency example, everything you
described with the agency example could be automated,
(37:41):
right. Every morning at 9:00 AM, your
agent wakes up and just starts browsing the Facebook ads
library. It signs more weight to ads that
have been running for the longest because that sends a
signal that they're more effective.
It recreates them. It finds the owner on LinkedIn
and messages them and message back.
(38:01):
It only creates the UGC ad once the owner has reached out and
said, yeah, sure, send me a freesample.
You know, like you can have to do all kinds of things.
Exactly. Yeah, I think that's how you can
scale services, have the repetition of not being
scalable. I think that's a good way to
scale them without having to hire an army of 100 people.
(38:23):
Yeah. I mean, anything scalable?
It's a question of how scalable is it?
How hard will it be to scale right?
Correct. That's how I.
Look at it. OK Romaine, this was perfect.
Where can we find you? I post content on LinkedIn.
I post content on X as well. Romaine Torres, founder of
Arcads. Feel free to reach out, ask
questions. I will try to look at the
(38:44):
comments as well. Happy to help.
OK, Arcads dot AI. Hey guys, if you're still
listening to this, it's probablybecause you haven't had a chance
to take your Airpods out. You're still mowing the lawn,
you're still driving, what have you.
If you're still here with me, I would really, really love and
appreciate a five star review onSpotify, Apple, or wherever you
get your podcast. It would mean a lot.
(39:05):
If you want to go the extra mile, share this episode with a
friend that might have an interest in starting a business.
It would mean a ton. Hope you have the best day of
your life today.