YT YouTube2Text

YouTube2Text → Transcripts → Watch me start & sell an AI service in 10 hours

Watch me start & sell an AI service in 10 hours — Transcript

by Nick Saraev · 20,738 words · 2,966 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Today, I'm going to start an AI
0:01automation service completely from
0:03scratch and build out all of the
0:04infrastructure that you need in order to
0:06get your very first paying client using
0:08this business model. I'm going to whip
0:10up the offers. I'm going to do the
0:11copywriting. I'm going to talk templates
0:13and blueprints. And I'm going to do it
0:14all live in front of you all the way up
0:16to getting my very first interested
0:18party on a call. And the reason why I'm
0:19doing this is cuz I just want to show
0:20you guys how accessible this business
0:22model is. So you guys have everything
0:23that you need. Basically like a copy
0:25paste playbook to get up and writing
0:26yourself. In a nutshell, this is what we
0:28are going to do today. We're going to
0:29start by finding and choosing a high
0:31demand AI service the clients are pretty
0:33desperate for right now. I'm then going
0:35to set up my lead scraping and outreach
0:36infrastructure in just a few minutes.
0:38Then I'm going to create an irresistible
0:40offer that practically sells itself. I'm
0:42going to generate qualified leads using
0:44two proven methods that cost very little
0:45in the grand scheme of things. And then
0:47finally, I'm going to generate my very
0:48first interested leads using these
0:50methods. So, doesn't need to be scary,
0:52doesn't need to be intimidating. I'm
0:53going to give you guys everything you
0:54need. Let's get started and make some
0:55freaking money. Okay, this is going to
0:57be the document. and I'll do everything
0:59off of. And this is the road map. I'm
1:01going to start by picking three niches
1:03and I'm going to determine two high
1:04demand AI services per niche to solve
1:07real client problems in them. I'll even
1:08show you how I do a little bit of
1:09research. After that, we're going to set
1:11up lead scraping and outreach
1:13infrastructure. This is a pain point for
1:14a lot of people that are watching this
1:16video. So, I'm looking forward to
1:16solving that. Then, we're going to
1:18create an irresistible offer using a
1:19simple copyrightiting formula that I've
1:20talked about, but I don't think a lot of
1:22people have seen. Then, we're going to
1:23actually generate real qualified leads.
1:25Then we'll chat in next steps how to
1:26actually, you know, like send a proposal
1:27and and close the deal and so on and so
1:29forth. Okay. So, you're going to see
1:31everything in this video from start to
1:33finish. I've named this watch me start
1:34and sell an ASA service in just X hours.
1:36I don't know what the X is yet cuz I
1:38decided to record it before I did all
1:39the stuff, which I think is the best way
1:40to do it. So, let's get up and running
1:42with just the very first step, which is
1:44picking three niches. So, I'm going to
1:46be relying on some of the templates in
1:47Maker School, which is my automation
1:49community, which basically just covers
1:51how to do all this stuff. So, the very
1:52first thing I'm going to do is go over
1:54to month one and then I'm just going to
1:55find some niches. Okay. So, what I do in
1:58day two is um I have a little document
2:00over here called the niche discovery
2:01spreadsheet. And what this document is
2:03is it's literally just a bunch of
2:05possible niches that you could pick for
2:07this business model. Then a bunch of
2:08different service lines that I could
2:10pick for this business model. Then I
2:11literally just mash them together to
2:12give you a ton of different positioning
2:14statements. And a positioning statement
2:15is just what I call something of the
2:17form I build X for Y. So X is the thing
2:20that we build and then Y is the business
2:22type that we are approaching. Okay. So I
2:24don't actually know what niches I'm
2:25going to pick yet. The reason why I
2:27haven't thought about this is because I
2:28wanted to put myself in as similar a
2:29position to as many of you as possible.
2:31So I'm just going to do a little bit of
2:32brainstorming out loud and maybe we'll
2:34cut it so it's not just me staring off
2:35into the void. When I'm thinking niches,
2:37I'm thinking digital businesses that I
2:40can source leads for very easily using
2:42automated methods. I'm also looking for
2:44businesses that have pain points that
2:46I'm pretty good at solving. And so
2:47historically, one of the main strengths
2:48that I have is like marketing, lead
2:50genen, growth for the most part. So I'm
2:52probably looking for businesses that I
2:54can realistically help with their
2:55growth. So there are a variety of
2:57different business types here that I
2:59think that I might have some prior
3:00experience with. I might have some sort
3:01of unique understanding of. And what
3:02I'll do here is I'll just I'll just pick
3:04a bunch. Hello to the anonymous panda
3:06that is currently on this page with me.
3:08So like I think website developers are
3:10okay. I'm also looking for niches that
3:11have money to spend, right? So I could
3:12do some sort of like website development
3:14agency. One niche that I've tackled
3:15before and I've had reasonable success
3:17with is the creative agency. And some
3:18creative agencies do website
3:19development. So, why don't I bold this?
3:21I could do sales with uh website
3:22developers actually, which is nice. Let
3:24me scroll through here and see more. Hm.
3:26IT consultants, e-learning consultants.
3:28H just trying to get some ideas here.
3:29You know, maybe I'll do video editing
3:31agencies as well cuz I've actually
3:32worked in video editing. So, I sort of
3:34have like some unique understanding of
3:35the problems that they face. And then h
3:37let's see here. I want to keep this
3:39pretty easy and pretty fast. I don't
3:41really want to spend too much time
3:42thinking about the niches. is I just
3:43want to get the 8020 in front of me and
3:44then I'll move forward. Probably do some
3:46content writing firms as well. No, I I
3:48don't think I want to do content writing
3:49firms cuz then I'd have two that I've
3:50run before. I've run a video editing
3:52agency and a content writing agency
3:53before. I don't want anybody to accuse
3:54me or I want to beat the allegations
3:55that I'm just doing businesses that I've
3:57already operated. So, let me think what
3:59else we got. Why don't we do real estate
4:01agents? A lot of people chatting about
4:03real estate agents now. That looks
4:04pretty good. Okay, so website
4:06development. I'll do video editing and
4:08then I'll do real estate. Okay, so I'm
4:09going to head back over here. I'm going
4:10to have my three niches laid out. We are
4:13going to do website agencies. So, I
4:14might just do creative to be honest. A
4:16creative agency might be a better way to
4:18conceptualize this, but we'll see. I'm
4:19just going to put this down for now and
4:21then we'll figure it out later. Then
4:22we're going to do video I'll just say
4:24video editors for now plus real estate.
4:27Okay, that seems pretty easy to me. We
4:29got three niches. We can actually
4:31already move on to the next step. And
4:33the next step is us going to be
4:35determining two high demand AI services
4:37per niche that solve real client
4:39problems. Okay, so this is another step
4:41where a lot of people get kind of caught
4:43up. They don't really know how to go
4:44about actually determining what services
4:46to sell to people, and that's totally
4:47okay. I don't have any like built-in
4:48resources for this, but what I'm going
4:49to do is just show you a simple
4:50framework that I always use when, you
4:52know, trying to figure this stuff out.
4:53So, assuming that I had no knowledge of
4:54website agencies and video editing and
4:56real estate. How would I go about
4:57actually getting it? Well, I would
4:59probably first turn to communities or
5:01Reddit threads or basically areas on the
5:04internet where a bunch of the people in
5:05these three niches congregate. So, what
5:07I'm going to do is I'm just going to
5:08find some communities where I've seen
5:10website agencies go in. Maybe like a
5:11website developer community or something
5:13like that. I'm going to join a bunch and
5:14then I'm just going to read top to
5:15bottom and see what some of the problems
5:16that they face are. So, I'm going to
5:17head over to
5:19school.com/discovery. School is just the
5:20main community platform right now. So,
5:22I'm just going to go here and then I'll
5:23type in website and let's see. Looks
5:25like getting tons of abs. So, I think
5:27the term website is probably not the
5:28term that I want to use. I imagine
5:30website is just like a general purpose
5:32term that like everybody has somewhere
5:33on their page and the search feature
5:34here is just pulling them up. like a
5:36website developer maybe. Okay. Software
5:37developer academy. Dev builders club.
5:40There's the AI automation agency hub.
5:42That's actually uh basically what we're
5:43talking about. So I'm not going to do
5:44that. Web agency accelerator. You see
5:46this? This looks perfect. This is
5:47exactly the sort of stuff that we want.
5:48Awesome. What else? Web designers
5:51collective. That's perfect. Exactly what
5:53we're looking
5:54for. Cyber value. No. Real estate
5:58developer. Not yet. Software engineering
6:00accelerator. Next level web design. Oh,
6:02this looks good. I'm looking for stuff
6:04that has to do with business, right?
6:05Like I don't just want a bunch of people
6:07building websites or whatnot. I want
6:08people selling stuff so that I could see
6:10what their problems are. Okay. But
6:11anyway, what do we got? So, it looks
6:13like one of these communities actually
6:15just allows us to read through all the
6:16posts without actually being part of it.
6:17So, what I'm going to do here is I'm
6:18just going to read through this and then
6:19I'll make my determinations afterwards.
6:21So, what am I looking for first and
6:22foremost? Well, first of all, actually,
6:23I am looking for wins post. So, it looks
6:25like there's a category called wins
6:26where this guy Tyler just sent 900 cold
6:27DMs in the last two days and closed two
6:29deals. That's pretty sweet. So, I like
6:30that. Looks like he's also using
6:31automation, which is very cool. Nice.
6:33There's a lot of questions about prices.
6:34What sort of price should I charge?
6:36Right? What platforms am I using? So,
6:38what I'm looking for really to be honest
6:39is I'm looking for I'm looking for
6:40questions. So, I see that there's a
6:42questions category. I'm just going to
6:43pump through here. And I'm looking for
6:44questions cuz I want to see what sort of
6:45questions that like the average website
6:47developer has. And this is just going to
6:48teach me more about the niche. I'm going
6:50to learn more about how to come up with
6:51a service that might effectively answer
6:53some of these problems. So, what's a big
6:54question I see? How do I find clients?
6:56Obviously. How to get clients?
6:57Obviously. How to get hired? Obviously.
7:00Agency versus freelancing. That's
7:01interesting. What sorts of clients
7:02should I pick? National, international,
7:04different platforms. How did you get
7:06your first client? That's big. How to
7:09acquire new clients. So, hopefully you
7:11guys are seeing that, you know, one of
7:12the main issues here is acquiring your
7:13new clients. And so, like a big problem
7:15statement first of all, and this
7:16honestly, you know, really got to be a
7:18rocket scientist to figure this out, but
7:19like some real client problems are
7:21basically all these niche are more
7:23clients. Okay? So, actually, let's frame
7:24this as a problem. Not enough clients.
7:28You know, we want more clients. That's
7:29problem number one across basically
7:31every niche. So that's a good problem
7:33that we could solve with a couple
7:34systems already that are coming to mind.
7:35But I'm not going to stop there. Let me
7:37do a little bit more research.
7:38Outsourcing got too many clients doesn't
7:40seem like a very common issue to me.
7:42Legal registration. Go highle website.
7:45People asking about specific design
7:46stuff. Very cool to niche down or not.
7:49SEO. SEO is cool. We could do like an
7:51SEO system. We could do some sort of
7:53like add-on for website developers. It's
7:54like, hey, you know, after you design a
7:56or build a website, you could also just
7:57like generate 100 blog posts for them
7:59completely free of charge or something
8:01like that. That's kind of cool. I like
8:02that. I like that idea a lot. I think
8:05that could have some merit, you know,
8:06like content generation for clients,
8:10right? Clients saying SEO. Yeah, really
8:12glad I found that. Was not expecting
8:14this at all. Let's just do website devs
8:16first and then I'll do the other ones.
8:17Okay, so obviously a little bit of
8:19business knowledge helps you do this
8:20faster, kind of understand the main
8:22problems that customers are suffering
8:23from. But I'll be honest, a lot of
8:25people have tried to automate this step.
8:26I think that automating the steps is
8:27probably the silliest thing that you
8:28could do. Like you actually
8:30understanding a customer problem is
8:31directly correlated with your success as
8:32a business owner. Like if I think about
8:34just the number one skill or handful of
8:36maybe top three skills that have made me
8:38a successful business owner, the biggest
8:39one is probably just understanding
8:41customer problems. Like reading through
8:43people bitching and moaning about their
8:45customer problems and stuff like that.
8:46So this is not something that I would
8:47automate. Like sure, maybe I'd ask chat
8:49GPT what are some problems businesses
8:50face? But I, you know, I wouldn't just
8:51take all that at face value. I'd
8:52actually go really deep. I'd actually
8:54like try and drink directly from the
8:55faucet if that makes sense as opposed to
8:57take some opinion or whatnot. Go and
8:59actually find problems that people are
9:00facing. So anyway, I found two, right?
9:01We found um not enough clients content
9:03generation for clients, SEO or whatever.
9:05That's good enough. I'm going to move on
9:06to the next niche, which is video
9:07editors. I'm just going to type in
9:08video. This is interesting. We'll open
9:10up that
9:11one. Scale with video. That's
9:13interesting. Educate with video. Hm. Get
9:15started with video. Maybe maybe we'll go
9:20editing. Editing lab. Editing layer.
9:23[Music]
9:25Edology. Short form training. Can I type
9:27in videography? How about that? That's
9:30cool. Wedding filmmaker. I guess you got
9:31to pay a,000 bucks for that. So, I think
9:33I'll leave that out for now. Wealthy
9:34filmmakers. Cool. That sounds good. All
9:36right. Let me just see how many of these
9:37I have direct access to right off the
9:39top. Looks like I only have one. Art FX.
9:40The rest of these I got to click a join
9:41button. So, just for the purposes of
9:43brevity, I'll just go through here and
9:44we'll start. So, everybody's talking
9:45about cold outreach again. That's cool.
9:48Looks like most of this community is
9:49this person Arthur just posting. So,
9:51this is not seeming like a very good
9:52community to start if I'm honest.
9:54Usually, you'll find that like some of
9:55these are just people's business models.
9:56They're like building a funnel. Okay, so
9:58some people are actually asking for
9:59videographers in specific areas. Okay,
10:01people are asking some pretty basic
10:02questions. People are advertising. I got
10:04fired from working with Gary Vee at
10:06Vayner Media. Wow, that's interesting.
10:07Okay, so color grading presets. That
10:08sounds like a pretty interesting
10:09problem. So, a lot of people want
10:10templates, out of the box templates and
10:12stuff, right? I wonder if I could solve
10:13that with automation. What are the
10:14simplest templates to get or whatnot?
10:16Hey, I'm a automation agency. I work
10:19specifically with video editors, and I
10:22wanted to give you a big list of all the
10:24best LUTs and templates and stuff right
10:25out of the box. That might work. I'm not
10:28entirely sure. People are posting their
10:30videos. I'm not seeing too many problems
10:32here. Okay, so I'm not seeing much
10:33success here. I'm just going to go to
10:34Reddit video editor subreddit. Let's try
10:37that. See, there's video editing here.
10:39There's editors, there's video editors,
10:40video editing requests. Let me just jump
10:42through this really quickly. See what we
10:44got. I'm really outing my uh myself
10:47here. LeBlanc
10:48[Laughter]
10:50mains.
10:51Yep. Been crushing some League. I'm just
10:54going to go to top really quickly. Just
10:56see top post for today. Alt tab all
10:58these. People are looking for specific
11:00help. This looks more technical.
11:01Probably not what I'm going for. Let's
11:03move on. This looks more like almost
11:05political. Ah, here we go. Video
11:07editors. This is like a smaller
11:09community about video editing. Longtime
11:10editor struggling to find freelance
11:12clients. Let's um open up this thread,
11:13read a little bit. Think I'm doing so
11:16much wrong. Cool. Let's see what this
11:18person's worried about. Obviously, there
11:21are a few people that are just talking
11:23shop and how do you make a living?
11:25That's that's a great question. Cool. We
11:27should be able to get some juice here.
11:28Help with my
11:30portfolio. Okay.
11:33H full-time position. Finding clients
11:35has never been my forte.
11:38It's a rap. This person's just talking
11:40about how tough it is. That's actually
11:42pretty good though because when you have
11:44really negative nancies like this, it a
11:45lot of the time like shows the fears
11:47that the niche is facing. Even TV, it's
11:49over because of automation. Well, that's
11:50unnecessary. I make a living with this.
11:53Yep. No jobs right now. Okay. Well, that
11:55one is obviously not enough clients.
11:57That's the main one. What you'll see is
11:58they're all kind of together, right? I
12:00mean, I already know that this is a
12:01major problem for real estate, so I'll
12:02just write not enough clients as well.
12:03Just insanely clear. All right. Well,
12:05yeah, that's all I got for video
12:06editing. If you joined all of these
12:08communities, kind of went really deep
12:10into them, I'm sure you'd find
12:11significantly more concrete problems.
12:13Let's now do real estate. Your real
12:15estate agents, global real estate
12:17agents, that looks great. Real estate,
12:18that's $999 a month. Probably not going
12:20to work for me. Let's um just filter
12:22this. So, for English, I don't think we
12:24can filter by free, can we? No. Okay.
12:27Another issue I'm seeing is busy work.
12:29When I say busy work here, I mean kind
12:31of mundane things like following up. A
12:33lot of the time, real estate agents,
12:34they make their money just by being in
12:36front of people enough like at the right
12:38times that when somebody or somebody
12:40that they know is interested in buying
12:41or selling a property or something, it's
12:42like, oh, you know, I just remember
12:43hearing about XYZ person 2 days ago. Let
12:45me send them a message. So, I think
12:47following up is probably a big problem
12:48point and just basic busy work and
12:50whatnot. I could probably come up with
12:51some systems to solve that. All right.
12:53So, as you guys can see, uh, am I like a
12:55rocket surgeon here, you know, niching
12:57super hard down into the specific issues
12:59that these guys are facing right off the
13:00get-go? No. I don't need to have the
13:02best problems right off the bat. What
13:04I'm going to do is I'm going to start
13:05talking to customers. I'll worry about
13:06figuring out like the nuanced natures of
13:08all these problems afterwards. Okay,
13:10we're just going to frontload the actual
13:11conversations with the business owners.
13:13We'll worry about everything after we're
13:14already up and running. So, this is
13:16sufficient for me. Okay. Now, after
13:17this, I'm going to have to come up with
13:18some high demand AI services. Now, you
13:20know, I have the problems. So, why don't
13:22I just list some services at the top of
13:23my head that I think could solve things.
13:25Okay. So, we'll have three categories.
13:27And I'm just going to aim for two
13:28services each. And they don't
13:29necessarily have to solve these problems
13:30one for one. We'll just come up with two
13:32each for three. So like let's just do
13:34website
13:35devs/creative agencies. All right. So
13:38the first major problem as we saw was
13:40not enough clients, right? So the not
13:42enough clients is the main problem.
13:44Well, what's the solution to not enough
13:45clients? Well, one of the simplest
13:47things that I sell all the time is cold
13:49outreach system. A variety of different
13:51cold outreach systems you could sell,
13:53but I don't even know what the shape of
13:54this is going to look like. I'm just
13:55going to get some cold outreach systems
13:57in front of my customers and I'll worry
13:59about exactly how to do it afterwards.
14:02Okay. But yeah, cold outreach systems
14:03are really big. When I say cold outreach
14:05systems, I just mean like scraping
14:06systems, email systems, systems that
14:08allow us to pump things into like a cold
14:10email platform. Basically, the same
14:12system I'm going to be building to get
14:13myself clients and then pitching that to
14:14other people. So, very meta, very much
14:16like using cold outreach to sell cold
14:18outreach for people, but it works really
14:19well. Next one is content generation for
14:21clients. So, I could absolutely see, you
14:22know, some instant content generation
14:24system for the websites that you just
14:26make for a client. Like, think about
14:27this. How cool it would be if it's like
14:29you already make websites as part of
14:30your whole thing. Imagine if you just
14:31had a single form or like a button that
14:33you could press or something that just
14:34immediately generated you 100 blog
14:36posts. And then you just took that added
14:37that to your client website and you
14:39said, "Hey, by the way, I already loaded
14:40this up with 100 blog posts for you,
14:41right? To simulate activity or whatnot."
14:43Very simple and easy add-on, which I
14:45think would uh probably be pretty cool
14:46and I'm interested in seeing how I how I
14:47might solve that. So, these look
14:48sufficient for me. I'm not going to
14:49think too hard on it. What we need to do
14:51after this is transform this into like
14:52an offer for cold email, but that's
14:53okay. Let's do the second niche now,
14:54which is videography. This might take me
14:56a little bit more time, but let's say
14:58right off the bat, I'm just going to
14:59sell cold outreach systems to these as
15:01well because we just saw how do we get
15:02clients? How do we get clients? We can't
15:03get clients. Clients don't exist. Uh,
15:05you know, this whole business model is
15:06dead. So, I probably sell some cold
15:07outreach systems to them, too. These
15:08cold outreach systems in particular will
15:10take a different approach than website
15:11devs, though. It's going to look a
15:12little bit different. I think we could
15:13probably do like a search intent system,
15:14which is like a job scraper, basically.
15:17So, you know, we'll scrape a bunch of
15:18jobs for people that are looking for a
15:19videographer to become in in-house and
15:21then we're just going to add them to
15:22some system and then we'll send them out
15:23to people that are hiring for that role.
15:25We're basically going to say, "Hey, I
15:26know you're looking for somebody to hire
15:27for X, Y, and Z role. I thought it'd be
15:28a little proactive, so I'm actually just
15:29going to send you a bunch of value right
15:31up front. I'm happy to do like a sample
15:32edit for you or something like that.
15:34100% free. Just send me over source
15:36material what you want and I'll whip it
15:37up for you within 24 hours." This would
15:38be a really good cold outreach system.
15:40I've seen a lot of people do something
15:41so that's actually how one of my video
15:42editors sold me. Okay, so what else are
15:44we going to do? Hm. Videography.
15:45Videography. videography. We could
15:47probably sell clipping. So, I see a lot
15:49of long form videos coming in. What we
15:51could do is we could offer videographers
15:53a way to very quickly and easily clip or
15:55maybe they upload the transcript of the
15:57video to my service and then my service
16:00automatically identifies timestamps.
16:02What else do video editors usually do?
16:03Hm. Exports. That's a big chunk of their
16:06workflow. Templates and whatnot. That's
16:08a big issue. I really like the idea of a
16:09clipping service for long form. Let's
16:11say a video editor is working with
16:13somebody for long form. You could very
16:14quickly and easily repurpose all that
16:15and make social media posts for them. So
16:17we'll either do clipping service. A
16:20clipping service is literally just
16:21something like Opus clip or V or
16:24Goldcast. What these are are you
16:26basically chunk in like a long video and
16:29then it just generates a bunch of short
16:30form ones. Same thing with V. Pump in a
16:32long video generates a bunch of short
16:34forms one ones. Same thing with Opus,
16:35right? Pump in a long one makes a bunch
16:37of clips like this automatically. So
16:39that's pretty cool. Um there's obviously
16:40a lot of value there. So I think we
16:42should probably do something like that.
16:43clipping service and or let's do like a
16:46content generation or maybe we make a
16:48bunch of Instagram posts or something
16:49about it. Okay, now let's do real
16:50estate. And what did we have? We had not
16:51enough clients. So, we could totally do
16:53some sort of email scaleup system. This
16:56is going to depend. Real estate's a
16:57little nuanced, right? There's like the
16:58buy side, there's the sell side. So, I
16:59might be selling a buy side system, but
17:01then that'll have like a fundamentally
17:02different meaning or purpose than a
17:03sellside system. If you think about it,
17:05a sellside system is basically like how
17:06do we get listings in front of as many
17:08people as possible. So, we could build a
17:09system that like, I don't know,
17:10automatically post it on a bunch of real
17:12estate agency websites or whatnot. Uh,
17:14that seems kind of lame to me, and I
17:15don't really think I could realistically
17:16do that in a short period of time. So,
17:18we could look for people to sell homes.
17:20Maybe what we do is we set up a pipeline
17:22for real estate agents that are looking
17:24to work with people on the sell side,
17:26and then all they do is they just supply
17:28the leads and then this system, you
17:30know, has really high response rates,
17:31and everything's like super high
17:32quality. I I think probably some sort of
17:34like email BTOC system would make sense.
17:37I don't know about sourcing the leads,
17:38but sending them out. And then I think
17:40we could probably do some sort of
17:42automated
17:44followup/reactivation system for them.
17:46Again, I'm just these are all off the
17:47top of my head. We're going to see how
17:48these evolve. But as a first pass, that
17:50looks pretty good to me. Do I know
17:52exactly how all this stuff's going to
17:53end up? Nope. I am just pulling some
17:54stuff out of my ass. I'm like pretty
17:56certain. I'm like 50 to 70% certain.
17:57Yeah, I think this might have some value
17:58for some people. That's enough for me.
18:00I'm just going to put it on the page,
18:01put together some offers, send it out to
18:02a bunch of people, and I'm going to hear
18:03their feedback first before I try and
18:05plan uh for every eventuality. Okay. All
18:07right. So, now that we have a bunch of
18:08these systems, what I'm going to do is
18:09I'm going to set up lead scraping and
18:10outreach infrastructure that I'm going
18:12to use to source people in all of these
18:13audiences. Then once I've done this, I'm
18:15actually going to go and I'm going to
18:16copyright and then start sending out or
18:17blasting out outreach. So, we're
18:19actually getting pretty far already and
18:20I think it's only been what, like 25 or
18:2230 minutes or something. Pretty simple,
18:23pretty straightforward. So, setting up
18:24lead scraping and outreach
18:25infrastructure. Well, there are variety
18:27of different ways to do this. Here's
18:28what I'm thinking of doing. Okay, just
18:30going to go all the way down here to
18:31make it a lot easier. What I'm thinking
18:33of doing is any sort of lead scraping or
18:35sourcing system is first we have to
18:36source leads. Then we have to scrape the
18:39leads. And then nowadays what we need to
18:40do is we also need to enrich the leads
18:42with some sort of personalization. Okay?
18:43So it's sort of like a three-step
18:44process. So what I'm going to do since
18:46I'm doing cold outbound is I'm going to
18:47start by sourcing leads. And there are a
18:49variety of different ways to source.
18:50Easiest one is definitely Apollo right
18:51now. And I don't know how long Apollo is
18:53going to work. So I'm just going to hop
18:54on that for sure. But you could also do
18:55like sales nav which is LinkedIn sales
18:57navigator. You could also do like job
18:59posts. That's pretty big. A lot of
19:00people doing job posts. After that, you
19:02scrape using a service usually like
19:03Amplify or some custom scraper you
19:05built. By the way, you could also do
19:06like a directory or something if you
19:08find a specific directory of people.
19:10Maybe a directory website. You could
19:11build your own custom scraper, scrape
19:13them, and your lead list would be way
19:14higher. But we're going to scrape these
19:15using Appify. And then at the end, I'm
19:17just going to enrich these. The way I'm
19:18going to enrich them is I'll probably
19:19like AI personalize, but I'm not going
19:20to AI personalize all of them. I'm
19:22actually going to just like try getting
19:23by with, you know, procedural variables
19:26and then AI personalization after. And
19:28then kind of the first step is we need
19:29to pump them into like infrastructure.
19:31we actually need to send, right? So, the
19:32main sender that I'm going to be using
19:33today is called Instantly. Instantly is
19:35just a very quick and easy way to get up
19:36and running with mailbox infrastructure.
19:38What I'm going to do is I'm going to
19:39source a bunch of leads, scrape them,
19:40and then enrich them first. That's kind
19:42of like, you know, what we're going to
19:42do now. Then in a few minutes, um after
19:44I'm done with that, I'm going to start
19:45uploading them to a platform instantly.
19:47And then after I'm done with that, I'll
19:49actually queue them up and we'll start
19:50sending the emails. Okay. All right. So,
19:51very first thing I'm going to do is I
19:53need to start sourcing leads. I see
19:54there are a variety of ways to do so.
19:56I'm probably just going to do Apollo cuz
19:57that seems pretty easy to me. Maybe I'll
19:58do job post, too. We'll see. But how
19:59would I go about this? I would actually
20:01just go Apollo.io. I'm just going to
20:03create three audiences here. First are
20:05going to be website development agencies
20:07or creative agencies. The second are
20:08going to be videographers and the third
20:10are going to be real estate agencies.
20:12You can see, you know, I work with
20:13creative agencies a lot, which is why I
20:14literally pumped in the search term.
20:15This is from the last time I ran a
20:16similar search. So, creative agencies
20:18sound pretty good to me. Um, maybe what
20:20we should do is we should have creative
20:21agency and then we could also add a term
20:23website here. Yeah, I could probably do
20:25that. I wonder if I type in creative
20:26agency and website. Maybe like website
20:29developer. Wonder if that would work.
20:31Maybe website agency cuz that's only 51
20:34people. Let's see. That's 54 people.
20:36H. Anything else that's limiting my
20:39search? How am I going to do this? I
20:41just type in website. We have 581, which
20:43is pretty cool. But as you see, a lot of
20:45these search results are like music
20:47website, outsource website, website
20:49factory. If I type in agency, what
20:52happens? If I type in agency, we have
20:5324,200.
20:55connective agency, MI6 agency, labeled
20:57agency. I wonder if we could add an
21:00additional
21:02filter. No, I don't want to generate an
21:05output for each record. Let me see here
21:08exactly what I would do. Keywords I
21:10would use website. So now I should be
21:13scraping agencies or co-founders,
21:16partners at agencies that include the
21:20term website. And so now I have multiple
21:22levels here. Now, what I'm going to do
21:24just to determine that this is pretty
21:25good, and we're never going to get a
21:26perfect scrape or a perfect data source.
21:28What I'm going to do to make sure it's
21:29pretty good is I'm just going to open up
21:31this page. Notice how many results here.
21:321 2 3 4 5 6 7 8 9 10 11 12. So, let me
21:36just zoom in a bit. 1 2 3 4 5 6 7 8 9
21:4010. Okay, so there are 10 results here.
21:41What I'm going to do is I'm just going
21:42to like do this two or three times. I'm
21:43going to go through the 10 results and
21:44I'm just going to see how many of these
21:46companies are actually companies I might
21:47want to work with with my search terms.
21:49I'm aiming for like 80%
21:51plus. So, let's open up all these and
21:54just see. Off the top of my head, you
21:55know, is this search good enough that
21:5780% of the leads I'm going to generate
21:58are probably companies I'm looking for.
22:00B TOC digital marketing agency whose
22:02name is Inspired. Okay, we do
22:05state-of-the-art websites. We built a
22:07bunch of websites. Okay, great. Digiv
22:08Vortex is good. So, we're one out of 10.
22:10This is Sydney website design agency.
22:12Good. We're two out of 10. H full
22:14service marketing agency. Okay, I think
22:15we could probably pitch that. That's
22:17three out of 10. Web design agency, four
22:19out of 10. Hm. I don't know if these
22:21people do websites specifically. It
22:22looks like they're more of a creative
22:23agency. They do storytelling. H I'm not
22:26entirely sure to be completely honest,
22:27but I think, you know, I think they're
22:30probably sufficient. Let me check their
22:32websites and see. Wow, this is very
22:34clean. Oh, they're in Calgary, too.
22:35That's uh where I live right now. Web
22:38site. Hm. Could I see maybe some of
22:40their work that they've
22:41done? Okay. Yeah. Yeah. Yeah. I'm seeing
22:45media production, business development.
22:48Okay. Branding. You know, branding
22:49probably has a lot to do with the
22:52website. Okay, they made a new website.
22:54Okay, so these guys do websites as well.
22:56Perfect. That's five out of
22:5910. Okay, this is like a brand agency
23:02tomorrow. I don't know if this is
23:03sufficient, but I'm just going to type
23:04in
23:05website. Oh, very clean. Very clean.
23:08Looking for the term website because I
23:09want to make sure that they make
23:10websites. If they don't make websites,
23:12this would not make any sense.
23:14Unfortunately, one thing that uh on a
23:16meta level, a lot of um website design
23:18agencies are doing is they're branding
23:20themselves as not website design
23:22agencies now. They're like, "We're
23:23multiddisciplinary creators. We're
23:25branding experts. We're architects of
23:29your vision, right? It's not like actual
23:31website." When I see stuff like rebuild
23:33our core business, though, odds are
23:34that's probably a website. So, I'm going
23:36to say yes to that
23:38one. Full service marketing agency. Uh,
23:41okay. Labeled agency. I feel like I I
23:44might have just doubled up on this and
23:46got two of these links, but I said that
23:47they were good. This is a website
23:48agency, web agency. Okay, so
23:50realistically, I'm thinking like nine
23:51out of 10 of my first search is fine.
23:5390% is pretty solid to me. I could go
23:55through more of this, but just off the
23:57top of my head, this is probably enough.
23:58So, I know that this lead list here,
24:00this source is probably sufficient.
24:01Okay, so very first thing I'm going to
24:03do, let me just make sure I'm nice and
24:05organized here. I'm going to go right
24:06over here. Then I will paste this in.
24:09This is just going to be my little
24:10scratch pad here. This is going to be my
24:12website website agencies URL. Okay, cuz
24:15the way you do this with Apollo is you
24:17just get the URLs. All right, now that
24:18we have the number one done, why don't
24:20we go number two? And this is good. We
24:21have 4,294 leads. I generally want like
24:23more than a couple thousand to run a
24:25test on. And we can get way more leads
24:26in this if we're smart. Uh, let me see.
24:28Let me see. Let me see what else are we
24:30going to do. So, we need videography,
24:32right? I wonder if I could just run the
24:33exact same hack here with the term
24:35videography instead. Now, odds are if we
24:37do this, probably going to get some
24:39duplicate records. Some of these website
24:40agencies are also going to do
24:41videography. Yes, I already see the
24:43tadpole agency. We already saw that
24:44labeled. So, already 20% of these are
24:46already the same agency, right? So,
24:48we're going to have some overlap, but
24:49videography is not the best term. Let me
24:51actually select this everywhere. So, h
24:55we do video. If I just do video, odds
24:59are they're going to do some
24:59videography, right? Okay. So, 4,188. And
25:02yeah, the labeled agency's with us as
25:04well. That's one duplicate.
25:07Tadpoles. That's another duplicate, but
25:08that's among 30. I think this is
25:09probably going to be fine. So, honestly,
25:11I'm just going to open up a bunch of
25:12these again. Four, five. Uh, yeah,
25:15labeled six, seven, eight, nine, 10.
25:17Double check. Animated explainer videos.
25:21This one is not in English. Videos.
25:24Okay.
25:25Videos. Videos. Cool. Yeah, these guys
25:29are the Calgary ones. I don't know about
25:31agency 360 video. H storytelling with
25:34video. Okay, that's good.
25:37Reasonable video. Cool. These guys do
25:40video, too. Video. Cool. They do video
25:42as well. Video. I'm not seeing anything
25:44here. So, let me just jump in. Double
25:46check that this is actually something
25:47that's produced. Man, what a clean
25:48website. I love these guys. Video. Hm.
25:51I'm not seeing any video here. If I type
25:53in
25:54portfolio, I see they have a show reel.
25:56They probably do videos. Yeah, I mean,
25:58if you Yeah, motion graphics. Looks like
26:00it. Okay, good. Miracle Video Agency.
26:02Again, this is in Italy. I'm not really
26:04targeting Americans here. I just wanted
26:05to keep this reasonably broad. Um, but
26:07yeah, video. Cool, cool, cool. So, this
26:09is another audience I'm going to pump
26:10in. We'll worry about the specifics of
26:12this later. So, I've mentioned we will
26:16now just do a
26:20videographers, videography agencies.
26:23Let's do that.
26:24Slashed. All right. And now we just have
26:26one more, which is real estate agents.
26:28So, let me just copy this, paste this
26:30down here because I want nice sizing.
26:31And then I'll just call this real
26:33estate agents. I don't want agencies.
26:36Agencies are big. So we'll just go real
26:38estate agent. I'm going to remove video
26:41here. And let's see how many of these
26:44are real estate agents. 1 2 3 4 5 6 7 8
26:509 10. Life improvement platform. Nope.
26:54Nope. Okay. No, this doesn't really look
26:56like good search. This is an example of
26:58one that is not good. It's probably
26:59because the job title, right? So, I'm
27:02thinking we're probably going to have to
27:03add like agent or maybe real estate
27:04agent. We could do like real estate
27:06broker or something like that. Well,
27:08real estate broker is not exactly the
27:09same, but should be sufficient. Okay, so
27:11as you see, we're getting way too many
27:12here. We're getting like 258,000. This
27:15is uh Greg W at Wallace Real Estate.
27:17This is a good sign for me because it's,
27:19you know, this sort of person we're
27:20going for like solo or or head owner of
27:23real estate brokerage. But yeah, we are
27:25getting way too many results to really
27:26do anything with this. So why don't I go
27:28to number of employees? Why don't we
27:29just look for people in the 1 to 10
27:31range and then maybe 11 to 20 range.
27:33This this will keep the companies pretty
27:34small. Um let me think. Are there other
27:36signals that we could use? Industry and
27:39keywords obviously need to be real
27:40estate, right? 41,000. Um we could chunk
27:44this down substantially further. Why
27:45don't I go location? We'll go United
27:47States because they tend to have more
27:48money and because we could just get this
27:5041,000 down to like 20 something,000.
27:52Okay. And yeah, still too many. So we're
27:54going to need
27:55fewer. Okay. Okay, I ended up filtering
27:57this down to about 16,000. This is still
27:59a little too many for me. And some of
28:01these work with like big corporations
28:02like Remax and so on and so forth. They
28:04actually have limited control and
28:05autonomy over the money that they could
28:07spend for marketing. Still, I don't
28:08really see this getting much better than
28:09that. So, I'm just going to call it here
28:11and put in my little real estate URL.
28:13Okay. All right. So, now we have three
28:15Apollo URLs. These all look pretty
28:16reasonable. What do we need to do with
28:18these? Well, we need to actually scrape
28:19these, right? It's one thing to have a
28:20URL, another thing to actually scrape
28:22these URLs. So, I'm going to pump all of
28:24these into Appify now. And I'm just
28:25going to run three manual runs. The one
28:27that I use is called Apollo code_crafter
28:30just because it's like a$120 per
28:32thousand leads. And uh when I say a$120
28:34per thousand leads, I don't actually
28:35mean that we are generating a,000 leads
28:36here with email addresses. We
28:38realistically generate like 2/3 of
28:39those. Okay. So let me see. Uh looks
28:42like we can only run this once at a
28:43time. So I'm just going to run this
28:45three separate times simultaneously. And
28:47I'm going to look for 2,000 leads first.
28:49Okay. So 2,000 for each. So just jumped
28:51on Appify, make an account. I'm at like
28:53a pretty high usage tier, but I think
28:55you can get by with substantially less
28:56than this. And then it's just $120 per
28:58thousand. So, not totally breaking the
28:59bank here. That's the first run. This
29:01going to be my videographer run. Let's
29:02go here and then save and start. Then
29:05I'm going to go to my real estate agents
29:06and then I'm just going to open up a new
29:08instance of this and then run it. Save
29:10and start. And now what we should have,
29:11if I go back to my actors, we should
29:13have three runs running simultaneously.
29:15Okay, so we have one here, one here, one
29:16here. Now we just need to wait until
29:18these finish basically because they're
29:202000s. Probably going to take us like 5
29:21minutes or so. Okay. And these are now
29:23just getting to completed. This one is
29:26600. This one is 1,300. So, I'm just
29:27going to set everything up for the first
29:282,000 run. And then as I'm doing it, the
29:30rest will finish. If I go to output
29:32here, I'm going go to export all 20,000
29:34results. There are a lot of fields.
29:35Okay. Tons of different fields. Um,
29:37realistically, and you guys are going to
29:39see that when I import all these fields,
29:40it's going to be crazy. There's going to
29:42be like a million in one. Uh, ideally,
29:44we wouldn't have that many, but um, it
29:46is what it is. I'm just going to
29:47download all these, show you guys what I
29:48mean. Uh, what I'm going to do is I'm
29:49going to have three. First of all, I
29:51don't actually remember what that
29:53specific export is. So, I'm going to
29:54export it now. Let's see. Going to
29:57append to current sheet import. I'm just
29:58using Google Sheets. It's the easiest
30:00way to do it. And we're going to start
30:02by doing all this
30:03manually. So, uh what is this? Yeah.
30:06Okay. So, this is the real estate. So,
30:07I'm just going to have one sheet. I'll
30:08call it actually I guess I could just do
30:10all this in one, right? Leads. I'll just
30:13call this leads. In fact, let's just go
30:15here for my own organizational purposes.
30:17Just go watch me start and sell. There
30:19you go. That'd be good. Here we'll go
30:21real estate. I'm just not going to use
30:24spaces. It'll make my life easier. And
30:26then what I want is I just want to get
30:27rid of all the fields that don't really
30:29mean anything to me. Okay, so first of
30:30all, just I want to count up the number
30:31of emails. There's a thousand in this
30:32list. So we actually only ended up about
30:3450%. Second of all, I want to delete all
30:36the fields that don't really mean
30:37anything to me. So realistically, what
30:40fields don't mean anything to me? Well,
30:41do you see how much space there is here?
30:43All these columns. Most of them are
30:45freaking empty as hell. What
30:47we want is we want to get rid of the
30:50vast majority of them. So I'm
30:53thinking there's two big swaths that I'm
30:55going to get rid of. Okay. The first is
30:59I'm going to get rid of everything after
31:01from
31:03AC all the way over
31:05to WH. Okay? And I'm going to do that
31:09for all three. And then I'm also going
31:10to prune a couple others. This way it'll
31:12just be a lot more easy and manageable
31:14for us. And you guys will see what I
31:15mean. If I'm to pump this into some sort
31:17of email enrichment service, I just kind
31:19of have to do this. So, what was it? AC
31:20to WH. Just going to copy and then
31:23delete. So, 500 columns are being
31:25deleted. Okay, this is a lot smaller
31:27now. A lot more manageable, right? Still
31:29pretty big, but it's a lot more
31:30manageable. Next thing I'm going to do
31:32is I'll go to personal emails. So, B I
31:34go to C. There you go. So, I think it
31:37was A C to WH and then BI to CE. Is that
31:42right? Hopefully that's right. Anyway,
31:44at the end of it, we want um like sheets
31:46that just all look the same because
31:47we're going to pump this into an
31:48icebreaker generator. We're basically
31:50going to have AI tell us a little bit
31:51about each person and then we'll use
31:52that to improve the the strength of the
31:54outreach. Okay, so this one is now done
31:56as well. Looks like it finished with
31:582002. So, I'm now going to download this
32:00same energy CSV. Go over here to sheet
32:032. Then I'm going to import. Go upload.
32:06There's a million in one ways to do
32:08this, by the way. Uh this is just how I
32:09personally like doing it. Then I'm going
32:11to append. That way it just goes into
32:13this one and then there are a lot of
32:14leads here and a lot of fields so might
32:16take a second. Okay, so it was so AC all
32:19the way to WH. It's going to be pretty
32:24far. This one has way more actually.
32:28Check this out. This one has a ton more.
32:32I think the reason why it works out like
32:34this is because basically LinkedIn where
32:36these guys get all their data from just
32:38stores all of the employment fields that
32:40you have. So yeah, this one actually
32:42isn't. We need to go even further than
32:44this. So never mind. You can't just copy
32:45the same uh things. I think you got to
32:47go all the way to here. So this one goes
32:49to ZS. Annoying. Anyway, um that's that.
32:52And then let me just go back to my
32:54poloscraper. Grab my last run, which is
32:56this one here. Go output. I'll export
32:59all these. Download these. And uh sorry,
33:02which one was this? Art
33:04director. Is this videography? I think
33:06this might be videography. Yeah, video
33:09agency. So, we'll just go video agency.
33:12Then this last one will be website
33:15agency and import
33:21upload. Okay. Now, I'm just going to do
33:23a little bit of reorganizing because I
33:24just want my fields arrayed linearly.
33:26I'm going to want the first name, then
33:27I'm going to want the last name, then
33:28I'm going to want like the company name,
33:29then I'm going to want a couple of other
33:31pieces of information. So, I'm just
33:32going to reset these so these all
33:33basically have the same information, the
33:35same columns. This is necessary for me
33:36to run the automation. I'll show you
33:38guys in a second. Okay, I have all these
33:40set up. Just for my own sanity, just
33:43going to make all of these fields about
33:45the same
33:46width. About boot. That's my Canadian
33:48coming out. This looks good. So, as you
33:51can see, we have first name, last name,
33:52headline, employment industry, industry,
33:53city, employment history, industry,
33:56city, country, email. Right? So, we have
33:57all of the fields that we will
33:59realistically need. And what I'm going
34:01to do now is I'm just going to pump all
34:03of this into a service. Well, my own
34:05service I'm going to whip up in a second
34:07to generate an icebreaker. So I'm just
34:09going to take all this information and
34:10I'm going to use it to generate, you
34:11know, some AI icebreaker for final
34:13confirmation check. How many emails do
34:14we have? We have about 1,100 for each.
34:17This is 1,026. Other was,233. This other
34:19one's whatever. So I'm just actually
34:21just going to run a,000 across all three
34:23of these. So 1,000 for real estate,000
34:25for video agencies, thousand for website
34:26agencies. And we're going to see where
34:27we land. Of note, I'm not verifying or
34:30validating my domains. So I'm not just
34:31using the verified ones here. You can,
34:33and I think in many cases you should,
34:36but I find that as a first go, right off
34:38the top, I just like to bust out as many
34:41emails as possible, not make any
34:43assumptions as to the quality of my
34:44data, and then after I run at least that
34:46first campaign, if something is like
34:47severely underperforming or if I do have
34:49high bounce rates or really piss poor
34:51deliverability, then I make my little
34:52changes afterwards. But for a first bet,
34:54I'm not even going to worry about that
34:55yet because I want this to be simple and
34:56uh nowhere near as complicated as I
34:58think most other people make it out to
34:59be. So now what I'm going to do is I'm
35:00going to add an additional column here
35:03called ice breaker. We'll just do column
35:05I for all of these. Okay, ice breaker.
35:07Then one more icebreaker. And I'm going
35:10to have AI go through and then generate
35:11ice breakers for all these. What I'll be
35:13using to make my life really simple is
35:14I'm going to use make.com which is a
35:16service that is basically no code. One
35:19of the foundational noode platforms
35:20similar to NAN, similar to Zapier and
35:23all these other ones. Realistically, if
35:24you're getting into AI and automation,
35:26you have no idea where to start. I
35:27recommend that you start with this. It's
35:29much easier and faster way of going
35:30about things. What we're going to want
35:31is we're going to want the sorry search
35:33rows. Go search rows to start. Then I'm
35:36just going to select the right account.
35:37I'm going to have to make this
35:38accessible to that account. So I'll go
35:40nick at leftclick.ai. Going to give
35:41myself editor access. No need to notify
35:43me. And the sheet I'm going to be
35:44looking for is watch me start and sell
35:46an AI service in just x hours. Right. So
35:49I'm going to go down to search method.
35:51Um using the wrong account here. One
35:54sec. Then I'm going to grab
35:55the ID up here at the top. Paste that
36:00in. It's saying that it's empty. I think
36:02just because of a bug. So let me retry
36:04this. The first one is going to be real
36:07state.
36:09Okay. Not entirely sure why I keep on
36:12getting empty for this. H might be a
36:14bug. For now, I'm just going to return
36:15five because I just want to make sure
36:17this actually works. So let me run this
36:18module. Get
36:20five. Cool. So it got the data.
36:22Basically, what I've done is I've just
36:23confirmed I can get the data from my
36:24real estate sheet here, uh, then pump it
36:26into make.com, which is good. After
36:29we've searched through it, we'll just
36:31call this icebreaker generation. Um,
36:34after we've searched through it, uh, we
36:36can now build out the rest of this flow,
36:37which is going to be pretty simple and
36:38pretty straightforward. We're basically
36:40just going to take all of the
36:41information in. Then we're going to, um,
36:44send a message over to GBT. Probably use
36:4740 for this one just for cost purposes.
36:49I'm going to have it generate ice
36:50breakers for all of these. So, let's go
36:53GPT4. Uh, let me just see actually
36:55what's the model prices. OpenI. Let's
36:57see the cheapest one right now. These
37:00are good, but they're quite expensive.
37:02This is 2 million. This is 40. This is
37:05one. Okay. Oh, wow. That's crazy. So,
37:08this one is affordable model balancing
37:10speed and intelligence. I'll probably
37:11use this one. GBT 4.1 mini, I think.
37:14Right. 2 million 1 million input tokens.
37:16I I don't know. We'll try it.
37:19one mini right over
37:24here. And then here is where I'm going
37:26to go and I'm going to uh design my
37:28prompt. So you're a helpful intelligent
37:30writing assistant. I always like to have
37:32a developer assistant prompt first. Then
37:34here is where I'm actually going to your
37:36task is to take is input a bunch of
37:40personal information about a prospect
37:42and then design a customized oneline
37:45icebreaker to begin the conversation.
37:48You'll return this icebreaker in JSON
37:51using this format. I'm going to say
37:54icebreaker and then rules write in a
37:58Spartan. How do you pronounce
38:01this lanic?
38:06Laconic. Oh, interesting. Renpartic
38:08lanic
38:09laconic tone of voice.
38:13Um, weave in uh context wherever
38:21possible. Here is a bunch of information
38:24about
38:26me, but you can make these ice breakers
38:28more
38:30personalized. Then I'm just going to
38:32list a bunch of stuff about me. And then
38:33down here, weave in context with my own
38:35personal information wherever
38:38possible. Keep things very short and
38:40follow
38:42the format. Hey,
38:45name we'll go new line. New
38:48line. Really respect X and love that
38:53you're doing
38:54Y. Thought I'd
38:57connect. Wanted to run something by you.
39:00That should be a good one. Okay. Keep
39:02things very short and follow the
39:04provided format. Leave in contact with
39:06me on personal information wherever
39:07possible.
39:09deniability possible. Try and imply that
39:12I like
39:14do believe. Oh. Oh no. Did I just delete
39:20the whole
39:21thing? That's
39:23brutal. Okay, I'm going to write the
39:25whole thing again. It's not going to be
39:26the same. That sucks. Be careful with
39:29the escape commands, ladies and
39:31gentlemen. That is tough. Do not want to
39:34escape. take as input a bunch of
39:37information about a prospect and then
39:39generate a customized oneline icebreaker
39:42to imply rest of my communicate is
39:45personalized. You'll return your ice
39:48breakers JSON format icebreaker and then
39:51here I'm going to say hey name love
39:55thing also a fan of other thing wanted
40:00to run something by you cool rolls right
40:03in a
40:04Spartan oh be very careful with that
40:07that's
40:09why keep things short and
40:11punchy short let's say imply familiarity
40:15wherever possible possible if you see an
40:17opportunity i.e. if you see an
40:19opportunity to imply that I like the
40:22same things believe the same things or
40:25want the
40:27same things as they do. Go for it. Sure.
40:30That should probably be okay. Let me
40:32save this this time. Learn my lesson.
40:34And I'm going to show advanced settings.
40:36And what I'm going to do next is I'm
40:37just going to do a bunch of examples for
40:39it. So we're going to go JSON object
40:40parse JSON response. Next up I will do
40:42user. I'm going to provide some um
40:44personalized information about the
40:46prospect. So, uh what am I going to do?
40:48We'll just we'll just spam it all like
40:50in this one,
40:52two, three, this. We'll do
40:56this. We'll do this. Okay. So, we're
41:00just going to add all this information
41:01in. And then next, the assistant's going
41:03to have some personalized stuff. But, uh
41:05we don't want this to be the la the we
41:07actually want this to be the last
41:08message before this. We actually want a
41:11couple other examples. So I actually
41:12just generate one. We are going to have
41:14an example of this exact format. Put it
41:17right here. I'll fill this in with some
41:19actual information. And then we'll have
41:20the icebreaker here. Then we'll just do
41:22this like twice. And then in this way
41:24we're going to have a model of what
41:25we're saying. So the value here is
41:27you're actually just reaching out to
41:28people as if you were doing totally
41:30custom outreach. So let me see. Is there
41:32any other information we can get about
41:33the Now that I'm thinking about it, I
41:34don't actually have too much information
41:35about the company, right? It's like a
41:37company oneliner or something.
41:38Otherwise, we're really going to be
41:39riding the plausible deniability train
41:41here. No organization description, huh?
41:43O, brutal. Yeah, you don't really get
41:44too much information. Oh, we got some
41:46personal emails, too. Interesting.
41:47Interesting. Huh. Well, actually, I
41:49didn't even realize we're getting so
41:50many personal emails. Instead of just
41:52using this email column, we should
41:53probably concatenate the two. I bet you
41:54if I did this, um, if I merge these two,
41:57might even get more. It's been a while
41:58since I've used the personal field here,
42:00but like I don't know if Apollo's gotten
42:01a little bit better at this over the
42:02course of the last little while or what,
42:03but that's crazy. Look at that. That's
42:05juicy. That is a very juicy column. I'm
42:07actually going to go back here now to
42:08where it says email. paste this in. Like
42:11check this. Oh man, there's a ton of
42:13emails here. Tons of emails. Okay. Well,
42:17I gotta say most of these are looking
42:20pretty good. If I go down to the very
42:22bottom of this uh 2000 list, if I go all
42:24the way up here now, we actually get
42:261500 emails instead of just and some of
42:28these are like hotmails and Gmails. Now,
42:30some of these aren't verified, so I'm
42:31probably going to have like much higher
42:32bounce rates, but I feel like I should
42:33probably concatenate these two now that
42:34I'm thinking about it into just one
42:36email column, right? Basically, if they
42:38double up, then we won't put the same
42:40info. Or maybe we will. Who knows? Okay,
42:42let's add one more column. We'll go
42:45emails. Okay, so I'm just going to
42:47filter this. Just sort A to Z. Should
42:50have all the emails now sorted. I don't
42:52like how this looks. So, let me just do
42:54some spreadsheet magic here. Um, I'm
42:56then going to cut this. Paste this in
42:59here. Then we can actually just resort
43:02this based off a toz. Now, we should
43:05have all of the emails, right? Yes, we
43:07should have some more emails here. Okay,
43:08cool. I'm going to do the same for all
43:10the rest of these just to get some of
43:11those um those bonus emails that I might
43:12have missed. I think that's
43:15important. Okay, so yeah, we just
43:17leveled up and got a few more email
43:19addresses, which is nice. Now that we
43:20have all that stuff done, we should be
43:22able to do our icebreaker. Um so why
43:24don't I just start with some leads here?
43:25I'll just do some manual ones. Let's
43:27see. Let's do H. So, a couple things.
43:29I'm seeing some duplicates here. So, we
43:32got to pre-process our thing. Uh
43:34actually, no, we don't. Sorry. instantly
43:35will automatically do this for us. But
43:37now that I'm sorting these manually like
43:38this, I could tell there's a fair number
43:40of duplicates in here. The reason why
43:42might be because we have different
43:43emails. So, let me just double check
43:46this. No, no, we can um dduplicate
43:53this. So, columns to analyze, the only
43:55one that I really want is H email. Just
44:00remove all instances of duplicates under
44:01H.
44:05So, looks like we had some duplicates.
44:07Might have just been because of how I
44:08processed that. Let me just double
44:10check. Same thing
44:17here. Okay, let's now replace this just
44:20with some false or sorry, real data
44:25here. Then we'll copy over the headline.
44:31Then we'll do the employment
44:35history. Then we'll do the
44:37industry. Then we'll do the
44:40city. Then we'll also do the country.
44:43Okay. So basically I'm going to feed in
44:45this. And then I want some personalized
44:47snippet as a result of this. So I wanted
44:49to say something like hey I know. The
44:51reason this is valuable is you can then
44:53make abbreviations as to the company. So
44:54instead of Maki agency we could just say
44:56Maki which makes it seem as if it's like
44:58custom written. We can also play around
44:59with like the capitalization and stuff.
45:01I'm not going to do that, but I'll say,
45:02"Hey, Anna, I'll add two new lines here
45:04so it jumps down." And I'll say, "Love
45:07what you're doing at Maki. Also doing
45:10some outsourcing,
45:12offshoring right now. Wanted to run
45:14something by
45:16you." Cool. That looks pretty good to
45:19me. What are we getting from here? We're
45:20getting it uh perceivably blunt with the
45:22love what you're doing at Maki. We're
45:23also getting, you know, if somebody were
45:25to scrape this, they would say
45:26something. I love what you're doing at
45:27insert company name here, but this just
45:29maki. It seems a little bit more
45:30personalized. And then these this is
45:33like a an exact variable outsourcing
45:35offshoring. So I don't know if I should
45:36really do this. So maybe I'll just say
45:39also doing some outsourcing right now.
45:41Wanted to run something by that looks
45:42pretty good to me. Okay. So we have one
45:44example right now. So what I'm going to
45:46do now, keeping in mind this is an exact
45:48duplicate of some data that's already in
45:50here. So, I'm just going to run this on
45:51three and then I'm going to see how good
45:53it is on three of my What is this? Is
45:56this real estate or what did I do?
45:58Websites
45:59or just double check real estate. Okay,
46:02let's run three and let's see how this
46:03goes. All right, we just finished the
46:06three icebreaker
46:09generations. Notice Blue Edge in Boon
46:11Beach. Been around South Florida real
46:12estate, too. Wanted to run something
46:13about that. Seems pretty realistic. Uh,
46:14it's not the worst. Okay, I don't really
46:17like that one. Okay. All right. Well, I
46:18mean like for the most part this is
46:19good, but it's not sticking to our
46:21formula. So, what I'm going to do is I'm
46:22going to go down reduce the temperature
46:23drastically. Just have it try and stick
46:25to the exact formula that I'm doing
46:27here.
46:29Okay. Use the the above
46:33format when constructing your ice
46:36breakers. Let's run this again. Let's
46:38see if this is a little bit better.
46:40Location isn't super important.
46:43just want
46:46the okay this is perceivable like I'm
46:49not selling real estate but this is I'm
46:51perceivably there's like a lot of heavy
46:53on the implication here I'm working in
46:54real estate around here right it's like
46:55well I'm not actually selling also a fan
46:57of finding the right fit okay well that
46:59kind of sucks it's pretty bad if I got
47:02this I'd be like h I mean real estate's
47:04a grind yeah know I don't like this
47:06we're going to have to give it a couple
47:07more examples unfortunate but it is what
47:09it is so I'm going to go over here uh
47:12Uh, let me just adjust the prompt a
47:14little
47:15bit. Maybe I'll just say make sure to
47:17use the above
47:19format. Let's tweak this one more time.
47:21Let's see if that is sufficient to get
47:23us a three out of three on the score. If
47:24so, I'll scale it up. Do 10. And then
47:26I'll
47:29check.
47:33Okay. Okay. No, no, no. One more thing.
47:37If you see the ability to make an
47:40acronym, if you see an opportunity to
47:43shorten the company name,
47:46say instead of XYZ
47:50agency, do so wherever
47:52possible. Say things like, "Love AMS
47:56instead of love
47:58AMS professional
48:01services." More examples.
48:12Okay, that looks pretty good. And now
48:14let's just add one more example. So
48:16we're going to add two, right? One's
48:18going to be user with the example. The
48:20other one's going to be assistant. I'm
48:21going to map the user back up
48:26here. Assistant, I'm going to map over
48:31here. Going to copy the assistant
48:33prompt. And now I'm just going to add
48:34one more user
48:36thing. Let's do this one
48:39here. This doesn't really give me too
48:41much information. So, let's maybe do
48:42this. Uh, let me just make sure the data
48:44is formatted
48:47right. Visionary agency leader. Oo,
48:50fancy. Wow.
48:53Okay. I found that stuff funny. That's
48:55all. Okay. Uh, information technology
48:57and services. And we have the location,
49:00Dubai, United Arab Emirates.
49:03Cool. And we'll say, "Hey, Adam, the
49:07human tech agency." So, I should go,
49:10"Hey,
49:14Adam, what you
49:17doing?" The human tech agency. The human
49:21tech. Let me just see if there's
49:25some No, this is Dubai,
49:30right? Let's take a peek at this. Oh,
49:33guys. Okay, this guy's actually pretty
49:36huge. I was
49:39uh kind of crapping all over it
49:42because it's visionary. Yeah. Wow. No,
49:45for sure. The human tech
49:53agency branding in particular
50:04is let's
50:10see just getting an agency off the
50:13ground wanted to run something by Okay,
50:15cool. That's
50:17plausible. Now, let's give this a go on
50:19the top three
50:21hits. Okay, first result was much
50:24better. Hey, Aaron, love Blue Edge. Also
50:26in real estate down here. Want to run
50:27something by you. That's super cool.
50:30We've now finished the other
50:31two. Don't like that it says elevation
50:34real estate. Also fan of the San Fran
50:35market. Wanted to run something by Yeah,
50:37that's not ideal, but it's honestly it's
50:38not the worst.
50:40So, what we're going to do is we're just
50:42going to force it to make acronyms or
50:44not acronyms, shorten the names a lot
50:51more. We'll say shorten the company name
50:54wherever possible. Say XYZ
50:58agency. XYZ
51:06agency. Do the same with locations. San
51:09Fran instead of San
51:12Francisco, BC
51:14instead of British Columbia etc. Cool.
51:19All right. Now that I've done this, what
51:20we need is we need one more to update a
51:22row. And what we're going to do is we're
51:23just going to go down this and then add
51:24icebreaker right over here. Okay. We're
51:26going to do this for real estate first,
51:27then we'll do this for websites, and
51:28then we'll do this for that other um
51:31Yeah. So, let me add the ID of this
51:34sheet right over here.
51:37Going to call this real estate. The row
51:40number I'm going to grab from
51:41this. Then the column range. We're just
51:44going to go A to
51:46Z. Okay. What do we want to update?
51:48Column I. The only thing we want to
51:50update with column I is the result
51:52icebreaker. Okay. So now I have these
51:54three. I'm going to give it one more
51:56test to see if it updates the column
51:58correctly. Cool. Looks fantastic. Now
52:01that this is done, I'm just going to add
52:02one more step here, which is a filter.
52:04I'm going to filter to make sure that
52:05the icebreaker column is
52:07empty. So, um only return rows where the
52:10icebreaker does not exist. Then I'm now
52:13going to go I don't know how many we
52:15realistically want to do. Well, I guess
52:17we just do all of them, right? So, I'm
52:19just going to go back here and then
52:21yeah, we're just going to run it. Um so,
52:23we're just going to try and go through
52:24all the way from the top to bottom of
52:25this whole sheet, which is going to be
52:26about a thousand records, which should
52:27take us realistically like I don't know,
52:29a few minutes. Just returned. Now we're
52:32pumping through one, two, three. Might
52:35take us a little bit longer than that
52:36actually. Thousand rows, maybe about 1
52:38second per row. Might take us uh yeah
52:41about like 15 minutes or so. So while
52:43this is running, what I can actually do
52:45so I could just copy this and run this
52:46three times. So I could just do it all
52:48simultaneously. Now I'm doing this
52:50manually. You really you do not have to
52:52be doing this manually. You can
52:54absolutely just automate this by pulling
52:56some system that um just does it all
52:59kind of sheet by sheet. you know,
53:00iterate all the sheets and then run it
53:02one row at a time. But I'm just on
53:03something of a time crunch today. So,
53:05obviously, I want to do this quickly.
53:07I'm just going to create three
53:09scenarios, one for each of these. One
53:11will be icebreaker videography. Another
53:13one will be website design or website
53:16agency. Then, I'm just going to copy
53:17these three. I'm actually going to run
53:19them all. Now, I can do this simply
53:21because I have a lot of token rate
53:24limit. I have a very high rate limit.
53:27Not a lot of other people can because
53:28their rate limits aren't as high as
53:29mine, but I can actually like feasibly
53:31run three of these. So, one
53:32simultaneously with the other two. Then
53:34I can just go through all of them
53:35generating the ice breakers. So, let's
53:37just double check that that looks good.
53:40Then we'll just go over here, double
53:41check that that looks good. Cool. And
53:43then I'm just going to make sure that
53:44all of these columns are good.
53:45Icebreaker. Icebreaker. Icebreaker.
53:47Nice. We should now just be able to send
53:49it. So, I'm going to do one. I'll do
53:51another. I'll do another one. And let's
53:54see how we're doing.
53:56Yep, these are all now running
53:57simultaneously. So, I should be able to
53:58do all 3,000 if you think about it, um,
54:01in about 15 minutes or so instead. All
54:03right, just grabbed a cup of coffee. I'm
54:05ready to go. While these are generating
54:08the rest of the ice breakers, which by
54:09the way, we could see if we just go to
54:10any one of these sheets. See, these are
54:12filling in as we go along. I'm doing the
54:14alphabetical order, which is why it's
54:16Dennis, Dennis, Derek, Tzita, so on and
54:17so forth. While all this stuff is
54:19happening, let's head over to Instantly
54:20and make these campaigns. I'm just going
54:21to call this one the title of this
54:23video. That looks good.
54:25Click continue. Now we have the
54:27Instantly campaign up and running.
54:28Instantly is the service we're using to
54:30send the cold emails just to make sure
54:31everybody's on the same page. Very quick
54:32and easy to get up and running with
54:34this. If I just show you guys the email
54:36accounts, you'll see that I purchased a
54:38bunch of pre-warmed email accounts using
54:40their pre-warmed inbox feature. The
54:42benefit to this is it just takes me zero
54:43time to get up and running. So instead
54:44of me having to spend two or 3 weeks
54:46warming up emails, I just clicked one
54:48button, spent a little bit of additional
54:50money, and then got myself these
54:52pre-warmed ones. They sell out of these
54:53pretty quick. So, if you guys want them,
54:55definitely just head over here, select
54:57one of them, and then uh you know, buy
54:59both the domain and then also get the uh
55:01email addresses here. We're seeing a
55:03trend where female names perform better
55:05than male names, which is why there's so
55:07many females like Britney, Kimberly,
55:08Allison, Caroline, Hazel, so on and so
55:10forth. So, if you see any of that and
55:11you're wondering, you know, what the
55:12hell I want a guy name, then don't sweat
55:14it. The way that you work with these,
55:15it's a different name, right? It's
55:16Audrey, it's Britney, it's Brooke, it's
55:18Carly, it's Elizabeth. The reason why
55:19it's all these different names is
55:21because instantly had to purchase the
55:23email inboxes ahead of time and warm
55:24them all up for us, right? So, a big
55:25issue that a lot of beginners think is,
55:27well, like, dude, this isn't me. This is
55:28Britney. This says Carly. That's all
55:30right. You have total plausible
55:31deniability when you send out the
55:32outreach. The important thing is not
55:34necessarily what the first name and last
55:36name on the email addresses. The
55:37important thing is like in practice, you
55:39will send outreach. People will be
55:40interested in wanting to work with you
55:42and then you get them on a call and then
55:44you'll show up on the call and they'll
55:46just assume that it's either like a
55:47misspelling of your name or it's
55:48somebody else in the organization or
55:49whatever. So you guys don't actually
55:50have to worry about that if you end up
55:51doing the pre-warm mailbox route like I
55:53did. But you know, you also don't have
55:54to do the pre-warm mailbox route. You
55:55could do something else. Anyway, let's
55:57actually spin this up. So I'm going to
55:58go to sequences now. Sort of have like a
56:00four-step copyrightiting, you know,
56:02structure for this and I'll run you guys
56:03through it all line by line. But I also
56:05want you guys to know that I'm going to
56:06pull some of the templates directly from
56:07Maker School and then I'm just going to
56:08make some slight edits to them because
56:10the way that I like to do things is I
56:11like to just operate off of templates
56:12wherever possible. Like I've already
56:14built the wheel a bunch of times. No
56:15need for me to rebuild the wheel. Just
56:16same thing with you. If templates exist
56:18that do 90% of what you want to do, just
56:20start with those templates and then you
56:21just 10x your leverage, right? You got
56:22to go from 90 to 100% as opposed to zero
56:25to 100. So, you know, now that we've set
56:27up our lead generation infrastructure,
56:29let's just do this and not knock over my
56:31cup of coffee. Time to create an
56:33irresistible offer using my copyrighting
56:35formula. Hell to the yeah. Let us go to
56:39maker school. I'm go down to my
56:42classroom and then I think the templates
56:44I'm looking for month four. Yeah, here
56:45we go. There's a bunch of instantly
56:47templates. These are the exact templates
56:49that I've used on a number of campaigns.
56:51You'll see that my uh my rates weren't
56:52tremendous on these. Like, you know,
56:54this one had a 4.8% reply rate still
56:56made me tons of money. This one down
56:57here was a little bit better, 11.6, but
57:00this wasn't directly as monetary, so it
57:01made me about the same. This one made me
57:036.1% reply rate. This one is um probably
57:05what I'm going to use for a couple of
57:06these. But what I'm going to do to start
57:07is I'm just going to copy over one of
57:09these campaigns. And then I'm going to
57:10worry about, you know, touching it up
57:11and making it look really sexy after.
57:13Okay, this is going to be one. And
57:14actually, I should probably create one
57:16campaign for each, right? So, I need to
57:17rename the campaign. So, what I'm going
57:19to do is I'm going to duplicate this one
57:20more time. This one will
57:23be real estate.
57:27Okay. This other one here, videography.
57:30And this last one here is going to be
57:33website. Okay, now I have three
57:36different campaigns. This is valuable
57:37because we're going to have three
57:38different audiences, right? And once I
57:39have that, now I can go through and
57:41actually I'll open up three tabs. One,
57:43two, three. Then I'm just going to copy
57:45this over. I'm going to go to sequences,
57:47paste, sequences, paste, sequences,
57:50paste. Cool. So now I have like a
57:52starter sequence. This is real estate.
57:53This one's videography. This one's
57:54website agency. So now I just need to,
57:56you know, change the language and
57:57whatnot. And I'm also just going to do a
57:58little bit of copyrightiting. So, when I
58:00designed this template, it was before
58:02ice breakers and personalization were a
58:03thing, which is why um we're seeing, you
58:06know, I found you on LinkedIn. I thought
58:07I'd reach out. So, realistically, this
58:09section over here, this is going to be
58:10replaced with the icebreaker. Back in
58:11the day, you didn't actually have the
58:13ability to do this, right? Or if you
58:15did, it was like very minimal or not
58:17very good. So, you used to have to imply
58:19a bunch of things. Nowadays, you don't
58:21actually have to imply as much. You
58:22could um get away with like some sort of
58:24AI personalization to make it seem like
58:25you've done some research on them, which
58:26they like. Anyway, in terms of the
58:28four-step copyrightiting process, the
58:29the very first thing that I like to do
58:30is I like to answer the question, um, is
58:32this spam? So, the quickest and easiest
58:35way to answer the question, is this
58:37spam? Is you just give them some
58:38personalization up front. That's what
58:39this first section is going to be. Okay.
58:41The second question is going to be like,
58:43okay, so who are you? If it's not spam,
58:46it's like, all right, we verified that.
58:48So, who the hell are you? And that's
58:49what this section over here answers. And
58:51this is just sort of putting yourself in
58:52the the mindset of the prospect, right?
58:55The third one is why does it matter to
58:57me? Okay, and why it matters is because
58:59well I'm not just some random guy. Okay,
59:01it matters because I made people a ton
59:02of money with this before. Then the
59:04fourth one is okay well then what
59:07next? Right? Like now that we verified
59:10all this, what do we what do we do after
59:11this? Well, uh what we do after this is
59:14we have some sort of offer or call to
59:15action. Offer and call to action. Okay,
59:18simple four-step copyrightiting formula.
59:20Don't feel like you have to copy my
59:22exact wording here or anything. As long
59:23as you just follow this structure, show
59:25them that it's not spam by personalizing
59:26it. Answer the who are you question and
59:29make it seem like it's you talking to
59:30them, not a faceless company talking to
59:32them. Justify why you're reaching out by
59:33basically saying, "Well, I made all this
59:34money for other people or I've added all
59:36this value to other companies." Then
59:37finally, what next down at the bottom,
59:39as long as you do this, you have four
59:40steps, you're you're pretty good, I
59:42would say, in most cases. So now, I'm
59:43just going to go through and I'm just
59:44going to make some adjustments to it.
59:45I'm going to throw in some of my own
59:47little spin and magic. And then after
59:49that, I'm going to split test this with
59:50another one of my sequences. Okay. Okay.
59:52Okay, so my real estate system is some
59:54sort of BTOC system. So I just need an
59:56offer for a B toC system with some sort
59:58of deliverable. So what I'll probably
1:00:00say is let's just make sure the
1:00:02icebreaker is good. So something like
1:00:04this. Hey Dennis, love Prada also work
1:00:06in real estate in St. George. Wanted to
1:00:07run something by you. Okay. Implication
1:00:09is I'm working in real estate. So it
1:00:10could be that I'm selling to real estate
1:00:12agents. It could be that I am a real
1:00:13estate agent myself. Whatever I'm doing,
1:00:15I'm just getting the click and that's
1:00:16all that matters. Now that we've gotten
1:00:17the click, I need to identify myself. So
1:00:19if I had something like this, what would
1:00:20my copy be? Really? wanted to run
1:00:22something by you. I'm just getting up
1:00:23and running. Okay, I'm just gonna play
1:00:25around with it and we'll figure
1:00:26something out. Okay, first campaign
1:00:29looks good. Hey, name love whatever.
1:00:31Also working in whatever and whatever.
1:00:33Wanted to run some bio. I'm new to the
1:00:34buy side. Put something interesting
1:00:35together a few months ago that works
1:00:36well. To make a long story short, this
1:00:38outreach system that uses AI to generate
1:00:39buyer leads around 30 per month on
1:00:40average cost just a few cents to run. I
1:00:42ran it for similar position at agent in
1:00:44SD. I'm leaving this um intentionally
1:00:46vague and we made 125K in a little over
1:00:48a week. This is real. I know this is
1:00:50completely out of left field, but I'm
1:00:51fairly confident we can do something
1:00:52similar. Just want to see if there's
1:00:53interest. You're one of the first agents
1:00:54I found looking into it. Would this be
1:00:55of value to it all? Wouldn't cost you
1:00:56anything. I'd pay for it all myself up
1:00:58front. If you're open, we could work out
1:00:59some sort of deal, but up to you. Let me
1:01:00know your thoughts when you have a sec.
1:01:01Thanks. Sending account name. So, yeah,
1:01:04a couple things. I mean, I don't know if
1:01:05this is going to work. We're going to
1:01:06try to I think it'll work. I've ran some
1:01:08campaigns like this before and it's
1:01:09worked. But, uh, the key in my humble
1:01:11opinion is to just make it seem like you
1:01:12are writing directly to the other
1:01:13person. You're not like putting them in
1:01:15a sequence. You wanted to make it seem
1:01:17very informal and like plausibly
1:01:19deniable and hey, I was just looking
1:01:22into your stuff and I wanted to run
1:01:24something by us, see if there might be
1:01:26interest, right? That's basically the
1:01:27vibe. Obviously, it helps when you have
1:01:28some social proof. So, I have social
1:01:29proof. I've run a very similar sort of
1:01:31campaign once for buyer agent and SD.
1:01:33So, that's pretty cool. Made a ton of
1:01:35money very, very quickly for them, not
1:01:37for me. I way underpriced. But yeah, you
1:01:39know, if you don't have something like
1:01:40this, then obviously that's one of the
1:01:41reasons why at the beginning when I said
1:01:42pick a niche, you know, do you have
1:01:43niches that you guys have some sort of
1:01:44experience and some sort of
1:01:46accomplishment, some sort of unique
1:01:47advantage. So yeah, I'm just going to
1:01:48copy this over and I'm going to do a
1:01:50very similar campaign for the other two.
1:01:51One of the reasons why I always
1:01:52recommend picking three niches is
1:01:53because once you've done the work for
1:01:54one, you just multiply the effort for
1:01:56the other two. So let me do something
1:01:57similar now for videography for a
1:01:59product and then I'll finally do one for
1:02:00website agencies as well. Okay, similar
1:02:02vibes. I actually used to run a
1:02:04videography company in Vancouver, BC,
1:02:06and I scaled it to about 10,000 bucks a
1:02:08month. Now, I didn't use this system to
1:02:10get all of that revenue. In fact, I I
1:02:12experimented with a similar system, but
1:02:14obviously it wasn't the same system. So,
1:02:15we're going heavy on the implication
1:02:16here. But, you know, again, we're using
1:02:18the plausible deniability. Hey, I put
1:02:19something interesting together a few
1:02:20months ago that works well. To make a
1:02:21long story short, it is blank, right?
1:02:23It's an outreach system uses and I find
1:02:25people hiring videographers agencies and
1:02:26then pitches them cost just a few cents
1:02:28to run. All of this is true. Obviously,
1:02:29when they work with me, it's not going
1:02:30to cost them a few cents cuz they're
1:02:31going to pay me for it. But the actual
1:02:33system costs a few cents. And then, you
1:02:34know, I'm kind of leveraging the listen,
1:02:35I know this is totally wild, but right,
1:02:38which I find some self-deprecation is
1:02:40pretty valuable. You know, it just makes
1:02:41things seem realer when you can insert
1:02:43like a sliver of almost doubt, right?
1:02:45Kind of clean. So, that's the
1:02:46videography one. That one took me like,
1:02:48you know, 30 extra seconds to put
1:02:49together. This is going to be the
1:02:50website agency. So, this is going to be
1:02:52really similar system, I think. Man,
1:02:54I've done the exact same thing for all
1:02:55of these niches, haven't I? Good god,
1:02:57I've worked in a lot of niches. I used
1:02:58to run a website design agency. It's
1:03:00like your classic digital marketing
1:03:02agency, but we did websites. So, I can I
1:03:03can get away with saying that, you know,
1:03:04I've run similar systems for myself,
1:03:06too. That's pretty funny. You know, I
1:03:08got to be honest, after you've worked
1:03:09with one agency, you've worked with a
1:03:11lot of them. You've worked with a lot of
1:03:13them. Very similar. Okay. Anyway, so
1:03:15please bear with me, but I put something
1:03:17interesting a few months ago that works
1:03:18well. To make a long story short, it's
1:03:19an outreach system that uses AI to find
1:03:21people hiring website devs, then pitches
1:03:25them around 30, then pitches them
1:03:35customized
1:03:40templates makes them the demo
1:03:44website run high converting. I ran
1:03:48something similar for myself back when I
1:03:51ran a site agency.
1:03:54This was in Vancouver,
1:03:57BC around five years ago and it got to
1:03:59the point where we're making well this
1:04:01is well over 10,000 bucks per month. No,
1:04:03this is completely out of left field but
1:04:04I'm fairly confident we could do
1:04:04something similar. Just want to see if
1:04:05there's interest. You're one of the
1:04:06first people I found when looking into
1:04:07it. Would this be a value at all?
1:04:08Wouldn't cost anything blah blah blah.
1:04:11Cool. That's sweet. Awesome. So, that's
1:04:14some campaign variant number one. What
1:04:16I'm going to do now is I'm going to add
1:04:17an additional variant. Okay. And this
1:04:20additional variant is now going to pitch
1:04:21like a fundamentally different thing,
1:04:23fundamentally different offer. And I'm
1:04:24just going to run all these side by
1:04:25side. So, what am I going to do for this
1:04:26one? I think I'm going to use this one
1:04:28here, which is this Longshot creative
1:04:31agency campaign. I really like this one.
1:04:33This one was really neat. And then I
1:04:35think for this one, I'm going to weave
1:04:36in um my own social proof, my own like
1:04:38YouTube channel cuz uh you know, it's
1:04:39obviously very big dick. If you could
1:04:41say I have a YouTube channel which has
1:04:42several tens of thousands of
1:04:44subscribers. People really like that.
1:04:45Okay. Follow-up reactivation system.
1:04:46That looks pretty cool. This is going to
1:04:49be the real estate one. So, we'll have
1:04:50some personalization, which uh in this
1:04:52case is going to be an icebreaker. Let
1:04:53me just make sure that's
1:04:55right. Checking in on my system here.
1:04:57We've actually now finished and wrapped
1:04:58up all of them. I think this one is just
1:05:00hanging. That might be a bug, but that's
1:05:02good. Let's now put some stuff in there.
1:05:05Okay. Personalization. It is a long
1:05:07shop, but I work specifically with real
1:05:09estate agencies. Basically, I make
1:05:13custom follow-up systems, AI follow-up
1:05:16systems that nurture old leads on both
1:05:19the buy side, sell side. I I run an
1:05:22Instagram channel that just hit 100,000.
1:05:25We're talking about stuff like this. I
1:05:27know this is a long shot, but I work
1:05:28specifically with real estate agencies.
1:05:29Basically, I make custom AF follow-up
1:05:31systems that nurture old leads on both
1:05:32the buy side, sell side. I run an IG
1:05:34channel that just hit 100K followers
1:05:36talking about stuff like this. There are
1:05:37a lot of agents picking this up right
1:05:40now. This will be my social proof right
1:05:43now. And I've personally helped a
1:05:47few
1:05:48scale their monthly sales. Oh, I don't
1:05:51know any off the top of my head. It's
1:05:53going to be tough. I'm probably gonna
1:05:54have to find that out later. Like go
1:05:56through maker school or make money with
1:05:57make and see what sort of numbers have
1:05:58helped people. But I imagine it's
1:06:00probably at least five figures. Anyway,
1:06:02am fully aware we haven't talked prior
1:06:05to this. You'll probably think this is
1:06:08cold outreach. I looked into I think
1:06:12I'll borrow a section from the previous
1:06:13one.
1:06:18outreach I looked into. To be frank, you
1:06:21were one of the first agents I found
1:06:23when looking into it and into it, but I
1:06:26think you'd be a good fit. Also, in my
1:06:29experience, client
1:06:31reactivation is super easy to
1:06:34implement, do this for you, totally
1:06:38upfront, at no cost. That make sense for
1:06:40you? If so, do you want me just go ahead
1:06:43show you an example or would you rather
1:06:44chat first? Kind of awkward over the
1:06:46phone. We can chat if you're game get on
1:06:47a
1:06:49call. Okay, cool. All right, I think
1:06:51this is plausible. I don't think this is
1:06:52going to be as good as the initial one,
1:06:54but what are you going to do, right?
1:06:55Let's just use first name Q is my
1:06:57question. We'll call it there. I don't
1:07:00think I could use PS if you're not sure
1:07:01who I am or a
1:07:03skeptical. Okay, anyway. Okay, let's
1:07:06just make sure I remember to fix that up
1:07:07later. Ooh, videography. Uh oh. What the
1:07:10hell am I going to do for videographers
1:07:11again? Let me think. Uh clipping. Oh,
1:07:14right. Clipping. Clipping. Yeah. Yeah,
1:07:15that's good.
1:07:25Basically, Basically, I build systems
1:07:27that
1:07:32automatically build AI systems that
1:07:34automatic long form videos like
1:07:37interviews, etc. to short clips for
1:07:41social media. I did this in my own
1:07:43agency once copy scaled revenue to 90k a
1:07:45month at around 50% margins. Oh, that's
1:07:47true actually. content like video
1:07:49interviews to short clips for social
1:07:53media costs sense to run but actually
1:07:56you know I could actually totally use my
1:07:57own case study
1:07:58here because I'm specifying long form
1:08:01content right does cost sense to run and
1:08:05I did do a similar system in one second
1:08:07copy that's pretty neat
1:08:10similar service business although
1:08:13obviously all about copyrightiting also
1:08:15is my own company so I had a lot of
1:08:16latitude but I want to see if I could
1:08:17build a similar
1:08:19thing for you is I think your vertical
1:08:21is one where you could easily do more
1:08:22than 2K if you have a good pipeline.
1:08:23Also in my experience, this is a really
1:08:26easy win to offer
1:08:28clients. Clients takes no extra work and
1:08:32you can add few hundred bucks in
1:08:37perceived
1:08:38value for
1:08:40free.
1:08:41Cool.
1:08:43Cool. Cool.
1:08:46Cool. All
1:08:48right, we'll also go sending account
1:08:50name. Very sweet. And then over
1:08:55here, let me think. Um, this website
1:08:57agency stuff, content generation systems
1:08:59for websites. I really like that angle.
1:09:01I really, really do. How can I make
1:09:02this? I mean, like I know I'm meta stuff
1:09:04here. These are pretty long. I normally
1:09:05don't make campaigns this long, but um,
1:09:07at the beginning, I just like split
1:09:08testing a variety of approaches. So,
1:09:10some I like to be really long and
1:09:12intense. Other ones I like to be really
1:09:13short and punchy. after I run the first
1:09:15last, you know, like where I actually
1:09:16run this on 100 emails a day probably or
1:09:18something like that per vertical. I tend
1:09:20to iterate pretty quickly after the
1:09:22first week and then I tend to just like
1:09:24test the next bout with like I'll
1:09:26literally grab an email, I'll cut it in
1:09:27half and then I'll just make it super
1:09:29punchy and I'll see if that one works
1:09:30better or not. So, I think what a lot of
1:09:31people are thinking is like, man, this
1:09:32is super long. How the heck is this
1:09:34going to work? Well, my tip for you is
1:09:35that once you understand the patterns,
1:09:37you can break the patterns a little bit,
1:09:38but you do have to understand what the
1:09:40patterns are. And that's sort of what
1:09:41I'm doing with the second pitch. Okay,
1:09:42so Q. And then this one here is going to
1:09:44be just like first name Q. Okay, so
1:09:48where the heck was I? Oh, and this is
1:09:50icebreaker, not personalization. I'm
1:09:52just forgetting all my variable names.
1:09:54Okay, so what do I do? I work
1:09:56specifically
1:09:58agencies. So wanted to run something by
1:10:00you. I know this is a long shot, but I
1:10:01work specifically with website design
1:10:02agencies. Basically, I built a system
1:10:06that lets you that upon creating a
1:10:08website for a client in whatever
1:10:11framework platform lets you
1:10:15generate around 100 blog posts for that
1:10:18client for free. I ran a similar system
1:10:20my own agency, Skilled Revenue 90K,
1:10:22though obviously all about
1:10:23copyrightiting also has my own company.
1:10:25But I wanted to see if I could build out
1:10:27you since in my experience is a really
1:10:29easy win to offer clients. takes no
1:10:30extra working to add a few hundred bucks
1:10:31in perceived value for free. Like you're
1:10:33going to deliver a website anyway.
1:10:35Website anyway. How much nicer would it
1:10:39look if it came preloaded with a bunch
1:10:41of blog
1:10:43posts? Anyway, um that looks pretty good
1:10:45to me. Cool. Just going to run it. And
1:10:48honestly, yeah, no idea how good these
1:10:50are. No clue. I'm just going to give it
1:10:52a go and we're going to see where we
1:10:54land.
1:10:55Let me just make sure that all these
1:10:56variables are right and then we can
1:10:58upload the leads to the campaign and
1:10:59start gunning it. Okay. Icebreaker.
1:11:00Icebreaker. Now this has to be
1:11:01icebreaker. Icebreaker. Cool. Okay. So,
1:11:05now that I have all this, I think I'm
1:11:06going to have to change some of these
1:11:07variables. Actually, now what I have to
1:11:09do is I have to add follow-ups. That's
1:11:11annoying. Um, okay. Probably grab
1:11:14follow-ups from an old campaign that
1:11:16I've already written. Uh, let's see this
1:11:19one here. Okay. Let's do this. TLDDR.
1:11:22Know this is a long shot but agents
1:11:25agencies think I can add a fair amount
1:11:27of value. See previous email. This isn't
1:11:31ideal. Ideally you'd restate your whole
1:11:33pitch but um to be honest my back is
1:11:35sort of hurting and I just want to get
1:11:36something up and running. So I'm
1:11:38probably just going to copy this. Use
1:11:39the previous subject and we'll just run
1:11:41one follow-up from now on. The value in
1:11:43a single follow-up is obviously you send
1:11:44more emails that way. There looks to be
1:11:47a new line here. So I'm going to add a
1:11:49new line here. Instead of re agents,
1:11:51it's going to be
1:11:52videographers, editing agencies. We'll
1:11:55say website creative
1:11:59agencies. Okay. Then because we're using
1:12:01the previous step subject, it should
1:12:02come nested with the previous one up.
1:12:05Then why don't we just do two days in
1:12:07between each. Okay. We'll do 1 2 3.
1:12:10Okay. And now we'll just go to my
1:12:11inboxes here. So I think I have 15
1:12:13mailboxes. 1 2 3 4 5 6 7 8 9 10 11 12 13
1:12:1714 15. Yeah, we do. So I'm going to do
1:12:19five each. So, one, two, three, four,
1:12:21five. This one. Uh, well, what am I
1:12:22talking about? Actually, let's just run
1:12:24all of them. Doesn't really matter,
1:12:25right? Yeah, we're just going to do all
1:12:26of them. 1 2 3 4 5 6 7 8. Looks good.
1:12:30And then we'll do all of them here as
1:12:32well. All right. Now, let's see what
1:12:34these settings are going to be. Are we
1:12:35going to do open tracking? Uh, no, I'm
1:12:37not going to. Delivery optimization,
1:12:39yes, please. Send first emails text
1:12:41only. Yes, please. We always send the
1:12:44first emails text only. Daily limit. Uh,
1:12:47well, I have 15 mailboxes times about 30
1:12:50a day is 450 daily limit. So, how many
1:12:52am I actually going to send per
1:12:53campaign? Let's divide that by three.
1:12:54We'll do 150. Send 150 here, 150 there,
1:12:58150 there. Easy. Show advanced options.
1:13:01Show advanced options. So, advanced
1:13:02options. Let's see what do we got. CRM,
1:13:05time gaps. Okay. Okay. Auto optimize.
1:13:08Yeah, we'll do reply
1:13:10rate. We'll do provider matching as
1:13:12well.
1:13:16So, provider matching and then reply
1:13:19rate. Then down here, I'll do the same
1:13:21thing. We'll do provider matching and
1:13:23then reply
1:13:25rate. And that looks pretty good to me.
1:13:28I think I'm just going to
1:13:30Yeah. So, now um we need leads, right?
1:13:33So, I don't actually have any leads in
1:13:34here. So, I'm just going to save all
1:13:35these first. I'm going to go um upload
1:13:37all the leads. And then I'm just going
1:13:38to double check before I make everything
1:13:40live. And then, um yeah, we should be
1:13:43good to go. So, let me just verify that
1:13:44we have all the ice breakers. Yep, looks
1:13:46like we do on that page. Real estate.
1:13:49Let's go down to video agencies now.
1:13:51Verify we have all the ice breakers
1:13:52here, too. No, not seeing any ones. Why
1:13:56not? So, do we get a bug or
1:13:59something? Okay. Well, how about the
1:14:01website aid? Oh,
1:14:04no. Oh, I ran them all. I just wasted
1:14:07like 3,000 ops. That sucks. I updated
1:14:10the same sheet. All right. Well, we'll
1:14:12just run them all again. No big deal. No
1:14:14problem at all. Sometimes you make
1:14:15mistakes and when you do, they cost you
1:14:18$6. And I just made myself a $6 mistake.
1:14:22video agency and then website agency.
1:14:25Rough
1:14:26stuff. But what are you going to do?
1:14:28Huh? Website
1:14:31agency. Don't do what I did. All right.
1:14:36Okay. All right. Let's run this again.
1:14:38So, I'm going to have to remove all of
1:14:40the old ice breakers because there's no
1:14:42Yeah, I think the reason we got that is
1:14:44because we didn't pass in a row number,
1:14:45right? There's no row number. Okay. So,
1:14:48let's run this
1:14:50again. Okay. Okay. Now, let's verify
1:14:53that looks good. It does.
1:14:56Sad. This one. How we
1:15:01looking? No, I don't seem to be filling
1:15:03that in.
1:15:06Am I doubling up again? I'm probably
1:15:09doubling up
1:15:11again. We're pulling from video agency.
1:15:14We're dumping in video agency. We're
1:15:16doing the row number. We're getting the
1:15:17Icebreaker. Oh. Oh, sorry. That's two.
1:15:20Icebreaker.
1:15:21Yeah. Yep. We got a lot
1:15:24of mistakes over here. This is also two
1:15:27icebreaker. So, Oh, in fact,
1:15:31it's even more. We'll go icebreaker
1:15:34there.
1:15:37And we just got to verify that this
1:15:38looks good. It does.
1:15:42Cool. All right. Now, let's make sure
1:15:44that everything is good. Website agency
1:15:46filling up. Yeah, website agency is
1:15:49filling up. How about this
1:15:53one? This one looks like it's filling
1:15:56up. Cool, cool, cool. And then how about
1:15:59this one? Yeah, looks good. All right,
1:16:01guys. So, do not waste 3,000 operations
1:16:03like me.
1:16:05You do not want to be wasting the
1:16:07operations. I wonder how many we've
1:16:09wasted now. Yeah, you can see my daily
1:16:11ops usage before like 500. Now it's at
1:16:13like 7,000. So, how many more I got?
1:16:15Okay, I have enough. I can run it
1:16:18all. Rough. I should note that you could
1:16:21build systems that optimize the hell out
1:16:22of this. Right now, we're sending 80
1:16:24Google Sheets calls and we're also doing
1:16:2680 OpenAI network requests. You could
1:16:28actually batch all this to one and you
1:16:29can make this one, well, you shouldn't
1:16:31do one, you should do two or three.
1:16:32usually like one every batch request for
1:16:34500 or so rows. But yeah, we could
1:16:35realistically like do this whole thing
1:16:37in like 10 ops, but I decided that I
1:16:39would rather go speedy and still show
1:16:40you guys how to get up and running as
1:16:42quickly as possible rather than worry
1:16:43too much about making it super perfect
1:16:45and then spending a bunch of time on the
1:16:46that really doesn't matter. You
1:16:48could spend money for time. And that's
1:16:49exactly what we're doing here. All
1:16:50right, just reran the icebreaker
1:16:52generator and now we have everything all
1:16:54good to go for all three groups. I'm now
1:16:57going to just download this as a CSV.
1:17:00So, we're going to do just the real
1:17:01estate one first, and then I'm going to
1:17:02go to my real estate campaign, which
1:17:04would be this one here. And then just
1:17:05going to upload the leads. So, head over
1:17:07to leads, add leads. Notice I'm doing
1:17:08this manually for now, but obviously we
1:17:10could automate the hell out of it later.
1:17:11What we need to do now is we need to
1:17:12match the column. So, first name, last
1:17:14name, headline. Nope. Nope. Nope. We
1:17:17don't need any of this stuff. Nope. Not
1:17:18location. No, no, no. Email, though,
1:17:20icebreaker. That's what we need. So,
1:17:22we'll go custom variable there. The rest
1:17:24of the stuff I don't really need either.
1:17:26Obviously, you can import as much of
1:17:28these as you want. I just didn't import
1:17:30many of them. That's a lot of fields,
1:17:31right? So, I do not do the verification.
1:17:34I am going to check for duplicates
1:17:35across campaigns list and the rest of
1:17:36the workspace cuz I don't want to be
1:17:38sending people to different campaigns.
1:17:40Usually, we end up with like a few
1:17:42people dduped. So, that's that. Then
1:17:44I'll do the same thing for the second
1:17:46campaign which was my video agency. So,
1:17:48let me go download as a CSV. Going to
1:17:50add some leads. CSV. Just drag and drop
1:17:53this puppy. Same energy, right? So, what
1:17:55we are going to upload is the
1:17:57icebreaker. The rest of the stuff is a
1:18:00no. And we have a lot of columns here.
1:18:02This should be good. We We've got the
1:18:04first name, I think, which is what
1:18:05matters. We may have some crossover
1:18:07there. While that's uploading, I'll do
1:18:09the last one. So, website agency. I'm
1:18:11going to download this
1:18:13CSV. I'm going to go to leads and then
1:18:16add leads
1:18:18CSV. Drag and drop. Same energy here.
1:18:21Just want to upgrade or import the
1:18:23icebreaker. Go all the way down to the
1:18:25bottom and then add all of these
1:18:26contacts. Okay.
1:18:28All right. So now that we've done this,
1:18:30what we have to do is we have to double
1:18:31or triple check that the campaign is
1:18:33good to go basically. So the way that I
1:18:35do it is with preview. So just making
1:18:37sure that all of the data here is filled
1:18:39in. There's no thank there's nothing
1:18:41under thanks. That's a problem. So why
1:18:44is there nothing under
1:18:46thanks? Sometimes there is no sending
1:18:49account name
1:18:50variable. So I go to variables here. I'm
1:18:53not seeing anything. Instantly sending
1:18:55account name. Let me just see if we
1:18:57could
1:18:58add some variables here. I always change
1:19:02the name a bit. Okay, here we go.
1:19:04Sending account first name. This should
1:19:06be good. I'm going to go back here. Go
1:19:08sending account first name. Oh, we have
1:19:10one extra variable over here. One sec.
1:19:13Okay, that looks good. That looks good.
1:19:15Awesome. So, I'm going to just copy this
1:19:17over because I would have screwed up
1:19:18otherwise. I'm going to save and then
1:19:21delete any additional ones I have. I got
1:19:24to do it all over, unfortunately. It
1:19:25kind of sucks, but it is what it is.
1:19:26Let's just make sure that everything's
1:19:27good. Okay. And then I'll do the same
1:19:29thing
1:19:30here. Also do the same thing down here.
1:19:34That and then I should Oh, uh, do I do
1:19:36that? Oh, actually, you know what? Looks
1:19:37like I already had it on the the
1:19:38follow-up. Okay. So, what am I going to
1:19:40do now? I'm literally going to preview
1:19:41this for all of them. That looks good.
1:19:43I'm going to go here. Preview. That
1:19:46looks good. I'm going to go over here.
1:19:47Preview. That looks good. Going to go
1:19:49back here. Preview. Uh, thanks again. I
1:19:52don't know why we didn't get the sending
1:19:53account first name. Okay. Yeah, that
1:19:55looks good. Save this. Preview. Cool.
1:19:58Looks good. This one here is
1:20:00preview. Cool. That looks good. Let's
1:20:02now go over here. Preview. This is very
1:20:05important. Got a preview. For whatever
1:20:08reason, we didn't have the sending
1:20:09account first name. So, I'm going to
1:20:10save, then preview. Okay, we got it.
1:20:12Going to preview this. Okay, we got it.
1:20:14This actually, I didn't really fully
1:20:16preview this. Let me not cheat. Okay,
1:20:18cool. All right. Now that we verified
1:20:20literally everything here, we can
1:20:21actually get up and running with this
1:20:22campaign. After this, I'm going to take
1:20:23a little break. Usually, it takes a day
1:20:25or so to start receiving replies. Also
1:20:27worth noting that it's technically uh
1:20:28Easter right now, so I'm probably not
1:20:30going to get a lot of people responding
1:20:31to this in the near term. Oh, sorry.
1:20:32There's one more thing I have to verify
1:20:33first. Schedule. So, I always like doing
1:20:35Monday to Saturday. I like doing 7:00
1:20:38a.m. to 6:00 p.m. ET. So, what I'm going
1:20:40to do here is go 7 6. Okay. I'm going to
1:20:43do Monday, Friday all the way to
1:20:45Saturday. Then I'll go 7 7 and then I'm
1:20:49just going to copy this and then paste
1:20:50it in two tabs. Okay. Okay. So now I
1:20:52have everything here. 7 to six should be
1:20:55good. I have double and quadruple and
1:20:57six tupal checked. Now I'm going to
1:20:59resume this campaign. I'm going to go
1:21:00back here. Now I have three campaigns
1:21:03that are up and running. Okay. So I'm
1:21:05going to go sign off for the day and
1:21:06then I'll check back in a couple and
1:21:08we'll see how the responses to this
1:21:10email campaign have gone. This is
1:21:11exactly what I would do. Starting a
1:21:13whole sequence from scratch. So we see
1:21:15here we've now checked that last create
1:21:17an irresistible offer. So we just have
1:21:19two things remaining. Okay, so I woke up
1:21:22at the crack of dawn this morning
1:21:23because I was very excited to see all of
1:21:24these email replies roll in. Uh, I got a
1:21:26couple of notifications over the course
1:21:27of the night. I decided I wasn't going
1:21:28to respond to anything just because I
1:21:30wanted to be able to review them all
1:21:31with you. But yeah, great news. In the
1:21:33last, I think 13 or 14 hours, we
1:21:35received a total of 11 replies. Seven of
1:21:38them were positive. So, this is one
1:21:39here. And you see even the maybe not the
1:21:42most positive of replies are still okay,
1:21:45right? So, Elizabeth, thank you for
1:21:46reaching out. I will pass, but
1:21:47appreciate you reaching out. Thank you.
1:21:48This fell says, "Natalie, thanks for
1:21:50reaching out. We currently subscribe to
1:21:51a platform that does what you're
1:21:52describing. It's called Y Lopo. I wish
1:21:53you all the best in this venture.
1:21:55Thanks." This person says, "No, thank
1:21:56you. I don't buy leads of any kind, and
1:21:57if I did, I wouldn't go after buyers."
1:21:59What's interesting is immediately after
1:22:00that, we just had a swarm of yeses. So,
1:22:03right after that, hey Samantha, thanks
1:22:05for reaching out. I'm open to work on
1:22:06Subnet if you could provide us leads for
1:22:07sure. Yes, I'd be interested. Let's
1:22:08meet. Let me know what time works for
1:22:10you. So, this person actually explicitly
1:22:11is like, "Hey, let's get going. Let's
1:22:13actually sit down and let's actually
1:22:14jump on a call." Obviously, something in
1:22:16my pitch, which in this case looks like
1:22:17it was the videographer pitch, uh was
1:22:19very um compelling to them. This person
1:22:21says, "Let's talk more." We also now
1:22:23have their phone number. We have all of
1:22:24their personal information. We give them
1:22:25a ring immediately if we wanted to. Um
1:22:27this person says, "Yes, I'd be
1:22:28interested. Send me more info." We're
1:22:30going to get a lot of like, "Send me
1:22:31more info uh replies to this sort of
1:22:33campaign just because the way that I put
1:22:34it together, right? I said, "Would this
1:22:36be a value of you? You know, let me know
1:22:38your thoughts and then I could put it in
1:22:40front of you." This more of like a value
1:22:41campaign. And when you have a value
1:22:42campaign, we're a bit lighter on like
1:22:44the let's book a sales call immediately.
1:22:45Obviously, we are asking, but not
1:22:47everybody will take us up on it. We got
1:22:48an I'm not interested from this person.
1:22:50We got a please send more details from
1:22:52this person. We got a thanks for
1:22:54reaching out. I'd love to test things
1:22:55out. I'm looking to expand my
1:22:56acquisition channels to the company as
1:22:57well. Mostly for your info. I'd like to
1:22:59get a bit more background on the process
1:23:00you have. So, I know we could work
1:23:01something out. Let me know if you can
1:23:02send a quick loom, but that's too much.
1:23:04Would love to just get a bit more info.
1:23:06This person here says, "Uh, hey,
1:23:08Elizabeth or Nick." So, I think the
1:23:09reason for that is because I said that I
1:23:11scaled 1 second copy. Yes. Saw the
1:23:13redirect to Nick's drive's personal
1:23:15brand. Is this a project under his
1:23:17umbrella? I'd love to see what it can
1:23:18do. Can you send me over a preview or a
1:23:19sample? I'm not going to say this is the
1:23:20best campaign ever. If we look at the
1:23:22stats, okay, we can already pull some
1:23:24numbers from it. The website agency
1:23:25campaign responded. We got five replies
1:23:28from it and every single one of those
1:23:29replies were positive. Okay. The
1:23:31videography campaign we got two replies
1:23:33from so far and 66% of the replies so
1:23:35far are positive. But we can already
1:23:37start making some determinations about
1:23:38like um where we go next and it's really
1:23:40only been a day which is pretty crazy.
1:23:41The real estate campaign has had three
1:23:43replies. Okay, so fewer replies than the
1:23:45website campaign by far and then we've
1:23:46had zero opportunities. So, we haven't
1:23:50really sent enough emails to be able to
1:23:51fully say whether or not this approach
1:23:52is going to work right now. But, I've
1:23:54run enough of these to know that when
1:23:56there's a a divide that's this big
1:23:58between a successful campaign, which
1:23:59basically has a 2% positive reply rate
1:24:01over here, then a campaign that has a 0%
1:24:03positive reply rate over here with, you
1:24:05know, 510 emails sent collectively
1:24:07across the two of them. You know, I have
1:24:09a feeling this real estate campaign
1:24:10probably just ain't it, dog. So, you
1:24:12know, a couple of minor optimizations we
1:24:14could do already if we wanted to
1:24:15significantly increase the number of
1:24:16positive replies we get. Like keep in
1:24:18mind that I I sent all this stuff on
1:24:19Easter Monday, right? So not most people
1:24:21aren't working on Easter Monday. I
1:24:22imagine the results would probably would
1:24:23have been a little bit better if I'd
1:24:24sent them during a weekday or work day
1:24:26or something. But a couple things that
1:24:28we could have done to make this even
1:24:29better. Uh a couple things we can do. If
1:24:31we cut this off, okay, then in an
1:24:34equivalent time period, so in I don't
1:24:35know 13 hours or however long it's been
1:24:37since I started this, then we would be
1:24:39able to take those 255 and we'd be able
1:24:40to disperse them across these two. So we
1:24:42would have been able to get another 125
1:24:44here, another 125 here. And given the
1:24:46reply rates, we probably would have
1:24:48received another three positive replies
1:24:50here. Maybe like another one or two
1:24:51positive replies here. So
1:24:52mathematically, like we could already be
1:24:54in the neighborhood of generating 10
1:24:55positive replies a day. As you see, a
1:24:57couple of those are legitimately like,
1:24:58"Hey, let's get on the phone right now.
1:24:59Hey, let's meet." If you can get them on
1:25:01a call, realistic rate is something like
1:25:0360 to 75% positive reply to a call,
1:25:05assuming that you have a pitch like
1:25:06mine. If you have a much more direct
1:25:08pitch, um, if you're like, "Let's talk.
1:25:10Let me put something in front of you.
1:25:11I'm going to charge you money for this.
1:25:12Here's my crazy superdetailed offer."
1:25:14and you just give them all the
1:25:15information up front. Obviously, you'll
1:25:16have a higher conversion rate on the
1:25:18back end to a meeting, but you'll have a
1:25:19lower kind of positive reply rate
1:25:20initially. So, we kind of play with
1:25:22those numbers. You know, there are a lot
1:25:23of other things that we could do. You
1:25:24know, if I go back to this Uni box here,
1:25:26we can blend together the booking a
1:25:28meeting side of things with just direct
1:25:30calls. This is something I'm seeing a
1:25:31lot of people do. So, you know how um
1:25:33one of these people back here had their
1:25:34phone numbers, right? And their phone
1:25:36number was like fully listed out. We
1:25:37actually have a lot of their phone
1:25:38numbers as well. Yeah, this person here
1:25:40just listed out their whole phone
1:25:41number. Um, oops. I don't actually want
1:25:43to open FaceTime. No, thank you. That'd
1:25:45be fun. Hey, what's going on, man? Yeah,
1:25:48just recording a video right now for
1:25:49several tens of thousands of people. How
1:25:51you doing? Uh, that might actually be
1:25:53pretty enjoyable that I'm thinking
1:25:54about. But, uh, this person just has
1:25:55their phone number, right? So, what you
1:25:57can do is instead of just cold calling a
1:25:58massive audience of people right off the
1:26:00top, like a lot of people think that
1:26:01they have to do. And I understand why
1:26:03you want to cold call. Cold calling's
1:26:04it's it's fun. It's nerve-wracking. It's
1:26:06enjoyable. It's like an adventure,
1:26:08right? But ultimately, in 2025, it's not
1:26:10very effective. Okay, so instead of just
1:26:12cold calling totally cold audiences,
1:26:13what you could do is you can run a
1:26:14pre-qualify campaign like this, blast it
1:26:17out to, you know, as many people as
1:26:18humanly possible, get a 2% positive
1:26:20reply rate, send 10,000 emails, get 200
1:26:22people saying, "Yeah, you know, I might
1:26:23I might actually want to do this." Then
1:26:25just call all of them. You know, the
1:26:27next time that you run a call campaign,
1:26:28every single one of those people will be
1:26:29warm and you'll be able to say, you'll
1:26:32not have to say, "Hey, you know, I've
1:26:33never talked to you before." It's like,
1:26:34"Hey, how's it going? You know, we just
1:26:35chatted over email. You just got back to
1:26:36me a few moments ago about um, you know,
1:26:38wanting to put a demo together. Just
1:26:39wanted to touch base. I think you guys
1:26:40are really cool and yeah, I just want to
1:26:41put myself out there. Like you can
1:26:42actually do those sorts of campaigns and
1:26:44you can do them really well. You know,
1:26:45when you've done enough lead genen, then
1:26:46you start to see patterns with this
1:26:47stuff and a big pattern in cold calling
1:26:49just for the people that are watching
1:26:50this video for your benefit is just most
1:26:52of the time when you call somebody,
1:26:54you're like instantly disqualified
1:26:55unless you have some previous
1:26:56connection. Quick and easy way to cut
1:26:57through all that noise. Some other
1:26:59things that I'd experiment with moving
1:27:00forward on these campaigns, uh, if I go
1:27:02back to this website one here, I would
1:27:04now experiment with cutting the campaign
1:27:06length pretty dramatically. So you could
1:27:07do this in the same campaign or you
1:27:09could do this by duplicating the
1:27:10campaign and then like comparing the
1:27:11stats one for one. Personally, I just
1:27:12like doing all the stuff in the same
1:27:13campaign cuz with the analytics tab,
1:27:15what we can do is we could actually see
1:27:16specifically which thing most people
1:27:18responded to. And it looks like so far
1:27:19basically everybody's responded to
1:27:21variant A. They haven't responded to
1:27:22variant B. So here you could actually
1:27:23just like add 10 different steps.
1:27:25Although keep in mind it'll split the
1:27:26volume across all of them. You could add
1:27:2810 different steps and then you could
1:27:29see, you know, which which step is the
1:27:31best. And maybe one of them is just like
1:27:34a quarter of the length. Maybe one of
1:27:35them is just like a quick little
1:27:36message. And maybe you could experiment
1:27:38with things like asset based outreach as
1:27:39well, which is a very big uh trend right
1:27:41now that usually generates you
1:27:43significantly higher positive reply
1:27:44rates with the drawback being, you know,
1:27:46you have to like generate an asset. So
1:27:47you can send fewer emails a day usually
1:27:49because you have less data by which you
1:27:50can generate an asset and it usually
1:27:52also costs you like a scent or something
1:27:53per lead that you send out because you
1:27:55have to, you know, run the token cost to
1:27:57generate the PDF that's customized or
1:27:58something like that, right? So yeah, a
1:28:00lot of stuff that you could do here.
1:28:01Hopefully you guys are seeing where I'm
1:28:03coming from. In terms of next steps
1:28:04responding to these particular leads,
1:28:06what I would recommend everybody here do
1:28:07is when they get to this point with
1:28:09their own email campaigns or whatever
1:28:10sort of outreach, have some templates
1:28:12that do 8020. So have some templates
1:28:14that like get you 80% of the way there
1:28:16that answer basic responses like for
1:28:18instance, let's see this one here, okay?
1:28:20Where the person says, "Let's talk
1:28:22more." Have a template that says, you
1:28:23know, "Hey, name, thanks so much for
1:28:26getting back to me. really enjoyed
1:28:28reading through whatever website they
1:28:30have or digital property they have and
1:28:32sure like let's get something on the
1:28:33books right now. Do you have a calendar
1:28:35that you could send over? Would you like
1:28:36me to do it? If not, I have three times
1:28:38available XYZ. Just send me over the
1:28:40closest one and I can give you a ring
1:28:41directly at this number or we could set
1:28:43up like a Google Meet. The point I'm
1:28:45making is try and minimize the amount of
1:28:46back and forth. If you want to make a
1:28:47campaign like this work really really
1:28:49well, what you do is you actually just
1:28:50like add the times directly in the
1:28:52outgoing email and you say like, "Hey, I
1:28:53can meet 7 a.m. on whatever date, 9:00
1:28:56a.m. on whatever date or 11:00 a.m. on
1:28:58whatever date." And then usually, you
1:28:59know, you'll get a lower positive reply
1:29:00rate, but you're going to have people
1:29:01give you the exact times that they want
1:29:03you to call them. And then you can
1:29:04actually just like do it without any
1:29:05back and forth. Um, but yeah, so what I
1:29:08would do is I'd make these templates,
1:29:09right? Once you have a couple of
1:29:10templates set up, so I don't know, like
1:29:12a positive reply template, maybe you
1:29:13have a template that's uh like, you
1:29:15know, please send me over more info. Oh,
1:29:16you have a template that explains
1:29:17something really briefly. Hook up the
1:29:19Unibox to your phone. Instantly has like
1:29:21a Uni box mobile app and so you can
1:29:23actually see the responses come in live
1:29:24during the day. Do whatever you need to
1:29:26do on your phone to prioritize those
1:29:27notifications and then have the template
1:29:29set up so that the second that an email
1:29:31comes in, you have like an internal
1:29:32rule. If it's between 7:00 a.m. to like
1:29:359:00 p.m. or something, your internal
1:29:36rule is I will respond to this email
1:29:38within 5 minutes. If you guys get leads
1:29:40like this, you need to treat them like
1:29:41gold. Okay? So, don't do what I did and
1:29:43wait, you know, so you could batch them
1:29:44all and have something nice and
1:29:45impressive to show on a YouTube video.
1:29:46The second you get a lead, that's when
1:29:48you you sit down and you message them
1:29:49back cuz leads are basically the closest
1:29:51thing you have to actual money at this
1:29:53point. And so, you need to treat them
1:29:54with a respect and reverence they
1:29:56deserve. Also, keeping in mind that it's
1:29:57cold outreach. You never really
1:29:59contacted a person. If you're not
1:30:00constantly on top of them, there's no
1:30:02reason for them really to get back to
1:30:03you. They have a bunch of other people
1:30:04in their email inbox trying to do the
1:30:05same thing. I'll be hopefully much worse
1:30:07than you are. Okay, great. Hopefully,
1:30:09I've shown you guys how you could whip
1:30:10up an AI service completely from scratch
1:30:12and get it to the selling stage. We're
1:30:14now at the point where we're validating
1:30:15the idea. We already have people that
1:30:17are interested in the loose thoughts
1:30:18that I put in front of them. What I'm
1:30:20going to do in subsequent videos is I'm
1:30:22actually going to work through building
1:30:23these services. So, building services
1:30:25for the campaigns that people respond
1:30:26positively to and showing you guys how
1:30:28to put together a specific templated or
1:30:30productized automation that you could
1:30:32then sell to audiences like this. This
1:30:34is the way you do business today. You
1:30:35don't build and then hope that people
1:30:37will come. What you do is you have some
1:30:38minimum viable product idea. Okay, you
1:30:40don't even have to put the product
1:30:41together. What you do is you start
1:30:43selling all the possible products you
1:30:44could put together. People respond
1:30:46positively to a few of them and then
1:30:47once you validated them, then you
1:30:48actually built the thing. So, this is
1:30:50going to be a series that shows you guys
1:30:51how to do all of that. And I'm really
1:30:52looking forward to putting it together.
1:30:54Thanks so much for making it to this
1:30:55point in the video. If you guys like
1:30:56this sort of stuff, you'll love Maker
1:30:57School. It's my 0 to1 daily
1:30:59accountability program where I basically
1:31:00show you how to do exactly what I just
1:31:01showed you in this video, but in way
1:31:03more detail. And instead of you just
1:31:04watching me do it, you actually get to
1:31:06do it with resources and templates and
1:31:08assets. Everything in the program is set
1:31:10up to make your life as simple as
1:31:11humanly possible. Whereas, if you guys
1:31:13already have a business that's thriving
1:31:14and you guys want to take it to the next
1:31:15level, perhaps by incorporating some of
1:31:16this outreach or maybe some other sort
1:31:19of standardized or templated system,
1:31:20definitely check out make
1:31:21moneywithmake.com. This is my premier
1:31:24and exclusive automation community
1:31:25currently capped at 500 where
1:31:27essentially we take something that is
1:31:28working then we sprinkle a little bit of
1:31:30automated systems productization
1:31:32templating and standardization to take
1:31:33you to the next level 25,000 bucks a
1:31:35month and beyond. Thank you again.
1:31:36Please like, subscribe, do all that fun
1:31:37YouTube stuff to bump me up to the top
1:31:38of the algo and I'll catch youall in the
1:31:40next video. Just

About this transcript

This page contains the full transcript of Watch me start & sell an AI service in 10 hours by Nick Saraev, generated from the public captions YouTube serves with the video. The transcript has 20,738 words across 2,966 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.