I Let Claude AI Get Me Clients for 30 Days (240 Meetings Booked)

I've built an AI outreach system that books between two and eight sales meetings every single day. There are no manually sending emails, there's no follow-ups, and there are no replies to keep on top of. The system finds leads, it writes personalized emails, and it sends them across five different channels. When someone replies, it handles the conversation and books them straight into the calendar. Now, the only thing my team does is turn up to the call. And in this video, I'm going to break down the entire system. I'm going to teach you how it works so you can build your own version of this as well. I've been in sales for over 10 years. I built my marketing agency affluent. co to eight figures in revenue, and I've sent millions of cold emails, and had hundreds of thousands of people watch videos like this on this channel where I've talked various different outreach systems. But in 10 years of doing this, a system like this has never been possible until right now. And I genuinely cannot wait to show you how it works. Right, let's get straight into it. So, this is Superbase. This is where we host all of the leads that we are reaching out to in this database. And so, 52,000 leads that we have fully enriched here, all of their details. And we even have lead scored each and every one of them, so we know the quality of every single lead based on a number of different variables. We've got all of our campaigns on the left, we've got our infrastructure as well. And so, this Superbase is a databasing tool where essentially storing the back-end infrastructure. There are actually three tools that we use. Smartlead is what we use to send out the emails. We have 300 warmed email accounts here. You see, we're probably going to blur a lot of information on the screen because I don't want Yeah, I don't want all these emails and the leads and so on getting links. But you get the idea, 300 emails warmed up for this campaign. And this is the actual back-end, this is the powerhouse of the system itself. And this is actually built on Cloud Code using VS Code, which is what you're seeing right now. Here we've got two instances of Cloud Code. What was I working on here? On the left-hand side, I lead scored a whole bunch of new leads on the system, and then on the right importing a new case study where we took an a business from 700k to 4 million in revenue because those case studies we actually use in our copywriting engine, which I'm going to come on to throughout this video. In fact, with that in mind, let's actually just get straight into

actually how I have built this thing out, starting with the machine itself, the back-end. So, the architecture is it was actually five components of this. We first of all have lead scraping, so the act of finding and scraping information on leads and then scoring them. We then have the infrastructure, and so how many domains and so on that we actually need. We have the AI copywriter, which is matching case studies to and copy and sending out those emails. And then we have surround sound, so actually sending emails, but also reaching out in a multitude of different ways like we have voicemail drops, we're also sending WhatsApps, and and something else which you're absolutely going to love. An AI appointment setting, so we are responding to every single lead that responds and booking them straight into the calendar. The whole thing is completely hands-off, and that is the machine. Now, I am going to break down each individual component, but quickly before I do, I'm going to run a competition on this video. I want to know what the single biggest thing is that's holding you back in your business right now. Is it finding leads? Is it building systems like this? Is it staying consistent? Comment it below underneath this video, follow me on Instagram @jordanplattin, and DM me the keyword AI outreach, and I'll send you the complete Notion doc that I'm about to go through in this video. And then, every one in 100 of you will also win a free coaching call with a senior member of my team. I'll announce the winners in the community post, so subscribe to check those out. Now, the reason I built this for my own agency is very simple. Outreach is a grind, okay? I think I just spat. Gross. You know outreach is a grind, I know it's a grind, definitely after all these years. You set these targets, maybe it's 10 calls a day, you start off with 10 emails, 10 DMs, and you do it for a week, maybe two, and then a client project comes in, or you've had a bad day on the phones, and then the outreach stops. Few weeks later, you're wondering why your pipeline is dry, you've got no sales. And I've seen the same pattern across thousands of agency owners that I've worked with over the years. It is the same story every single time. And everyone knows the strategy now, right? It's very simple. We find people who we need or who need what we have to sell, we send them a message, and then we follow up. But actually doing that consistently every single day when you've got clients to manage and fires to put out, that's where things start to fall apart because humans are inconsistent. And the more manual the outreach is, the more it depends on you actually having a good day. And so, — I suppose I just took the human out of it. So, let me run you through the maths. Once you see the numbers on the back-end, the rest of this video is going to make a lot more sense, you'll get a lot more taste for it as well. So, let's break it down starting off with

the sending volume. So, we start off with our TAM, our total addressable market. This is every single company in our niche. What I'm going to go through right now is how we determine how many outreaches per day we're going to send. Most people say, "Jordan, how many should I send? " Like they've just got to pick a number. No, the number of outreaches we send on a daily basis is determined by the maths that we're going to go through right now. So, get your total addressable market, every single company in your niche. I don't care what you sell, you sell it to a specific kind of person, write that down. Then we have our serviceable market, our SAM. These are the ones we actually want. This is the ICP, the ideal customer profile, the ones that are qualified. We then have our enriched SAM. So, out of all of those people in the SAM, how many people can we actually find verified decision-maker emails for? And then, what we do is we take that enriched SAM number, we divide it by three, which is three months, and that is our daily sending volume. Why three months? Well, we can rotate through every lead every three months because they forgot about us until we put them back in the system to squeeze a yes or a no out of them. Now, what this translates into for my business specifically is 880 new leads per day entering this system, 3,000 personalized emails a day with roughly three follow-ups, and between 0. 2 and 1% meeting book rate on leads. That is the total amount of people that are booking meeting sales meetings based on the total amount of leads, two to eight meetings per day, 60 to 240 meetings a month. But of course, not everyone will show up. On cold outreach specifically, I benchmark at 60% show-up rate. That's 36 to 144 people showing up to a meeting. That's 7 to 29 new clients signed every single month on a very, very conservative 20% close rate on attendance. We actually close 40% across the business by standard, but we also get a lot of inbound. Now, for us, we have an average lifetime value on the smaller leads that we reach out to of 60,000 pounds. That is the total amount of money they spend with us over their lifetime. So, it's 420 to 870,000 pounds in lifetime value every single month from a fully automated system. That is insane. Manual outreach used to be free to 30 to 50 new leads every single day. That's all one person could handle. We used to hire multiple people in order to do this. This system that handles 880 leads a day, although actually, the total amount of leads that you can reach out to is dictated by your total SAM, okay? So, you might have 3,000 leads a day that you can reach out to. Great. All you need to do is just increase the back-end infrastructure, of which I'm going to run you through in this video. And this outreach system is — doing this every single day without excuses, 24/7, okay? And we don't have to have all these human errors and issues and so on. And we can ensure the absolute quality. What changes The what is the difference between a 0. 2 and 1% meeting book rate, okay? Well, it's the quality of leads we put into the system. And so, using a AI system, which I'm going to teach you, we can ensure that only the highest quality leads come in. But it's also the quality of the outreach itself. It's hyper-personalization. It's making sure that every single person feels that this email, this outreach has only been sent to them. Now, we weren't able to do this previously just by importing the company name and the first name and so on. We're able to write entirely personalized emails now, which means that we might even get cold email. If you get personalization right, if you match a relevant case study to every single lead, if you follow up across different channels, those numbers go up significantly. But even at the most conservative end, this maths is ridiculous. And that's This is what really I suppose what most people don't get right about outreach. The volume matters, obviously, that's what most people have, but volume combined with personalization and a multi-channel approach, that is a completely different ball game. And that's what will separate your outreach from other hundreds of other generic cold emails sitting in your prospects' inboxes. So, let me show you how to actually build this machine. We'll start off with from the top. This is going to be building your hit list

and with a waterfall, okay? And a waterfall is just a sequence that we follow through scraping leads until we go from our total list to a verified list, and we want to ensure we have, obviously, the highest percentage of people that we actually have verified. So, the waterfall goes as follows. Number one, we define our ICP. So, what industry, what revenue level, how many employees, the geography, and tech signals. You will actually use a variety of different ones of these in order to build your initial list. You're going to scrape the companies through database providers matching our criteria. There'll be various different databases that'll be relevant for you based on your industry. Cloud Code will be able to help you identify these. We're then going to find decision-makers. It's going to be the founder or CEO if you're a smaller agency, but if you're working with big companies like us, then CMO, head of marketing, broadly speaking. We're then going to find verified emails. So, we start off with a primary provider, so wherever we scrape the rest of this information from, but then we want to create multitude of fallbacks, okay? So, one that you could use would be Lead Magic, for example, and then you might use a guessing logic with hunter. io. By guessing logic, I mean, if you the name is Jordan Platten and it's Jordan it's @affluent. co is the domain, then you might be searching j. platten, jordan. platten, j platten, and so on and so forth. And you run those through AI, which again we weren't able to do without AI. We're now able to run these sophisticated guessing logics to be able to pull many more emails out of the system. And then we score every single lead on a scale of 0 to 100 and we store that in the database with the leads. That will change based on your criteria. For us, we know a certain level of products is going to be a good indicator. We know what kind of ethics the brand has. industry they are in, that sub-industry, what personas they're hitting. These are all going to be indicators for whether or not we're going to be able to get them maximum return on investment and therefore they're going to pay us the majority money and have the highest lifetime value. And then finally, we find mobile phones only for the top leads for multi-channel approach, okay? And we will run through that. Why do we do only for the top leads? Finding mobile phones is probably five times more expensive than finding email addresses. So, we want to be conservative. We don't just want to waste our money on low-quality leads when we are scraping those mobiles. Let me put some real numbers on this. I scraped about 76,000 companies that match my ICP. If I only used one enrichment provider, for example, my favorite is clay. com, I probably would have ended up with about 21,000 contactable leads. Maybe a little bit more on clay because it already has built-in waterfalls. But now with AI, we can build completely custom waterfalls. And so in this instance, now if I just used one enrichment provider, I would have ended up with probably 21,000 contactable leads give or take based on just using one different software, which sounds all right, but there's still 55,000 leads just sitting there with no email address. And so what the waterfall does is it takes those failed leads and it runs them through different providers. And each provider has access to different data, right? So, one misses it and the next one catches it. And then by the end of it, we had 52,000 contactable leads like you saw on Superbase at the start. That is triple what a single service provider would have given me. And then the custom scoring pieces, it's massive. Not every lead is worth the same investment. Some leads just get an email. The best leads get emailed, they get WhatsApp, they get a voicemail, and we even do a handwritten letter on their desk. So, scoring just means we don't waste money on leads. Right, let's get into some of the back end, the infrastructure equation, the

formulas first of all. So, our daily sends is our enriched send divided by 90 days. We've already determined that. Now, what do we need on the back end as far as email accounts are concerned? This has changed a lot over time, so make sure you follow this religiously. Email accounts, the total amount we need is our daily sends divided by 10. So, we're sending 10 emails per email account. Then as far as domains, we're going to divide email accounts by 50 or 15. So, we have 15 accounts per domain, okay? And so that gives us the total amount of domains we actually need to buy, which in my instance was around 20 domains that I bought initially. You're then going to warm up those emails for two to four weeks before sending live. Two weeks is the minimum. Four weeks is you don't really need much more than four weeks, but obviously the more the better. Warming up is just a process of a sending tool, okay? So, we have SmartLead here. Basically simulating a really good email health. It'll be sending emails, it'll be responding to emails, and basically going back and maybe I can click on one of these without kind of giving away the address. Yeah, you can see in the last 7 days with sent 55 warm-up emails there. This is how many it's received and it's 0% safe from spam, so none ended up in the spam. 100% of the warm-up emails landing in inbox, so it's really healthy on that domain at the moment. And so we need to make sure that we do that. That is there's it's out of the question skipping that step. So, the rules, never send from your main business domain. If you're contacting like the email address that you contact clients on, never use that. You could really hurt it. Make sure you've got SPF, DKIM, and DMARC on every single domain. You can set this up very easily with Cold Code now, although you could just pay someone on Fiverr to do it for you as well, which is what I do. Always have warmed-up backup accounts ready to go. There can be issues where you get fall into spam. People will report you from time to time. And so have like 10% generally is the rule that I like to keep on the back end. Now, no one wants to talk about this part because it's boring, okay? People don't like it, but this matters more than everything else I've shown you so far. You can have the best results in the world, but if they are landing in spam, nobody is reading them. And spam filters have only gotten stricter over the past year. And it's not even hard to set this up, it's just tedious. It takes a bit of time. No one wants to sit there and do it, which is why I pay people on Fiverr to do it. But if it's not an option for you to pay someone right now, then you can just follow the instructions of everything I just went through. Although it's very cheap to get someone to put do the infrastructure. And this system for me is probably costing $2,000 a month at the volume which we're doing, which is outrageously cheap in comparison to the LTV. But of course, it can be much cheaper than that in like the low hundreds depending on what kind of volume you are sending. So, once you've got this infrastructure dialed in, you need something to actually send, right? And this is where it's starting

to get interesting. This is the AI piece. So, let's go into the AI copywriter. Now, there are three inputs. The first input is our actual copywriting framework. There are a bunch of different frameworks that we teach. I've got other videos on this channel as well going through email copy. I'll put some links in the description, you can go ahead and watch them. Probably the most common is AIDA. And so what we do is we grab their attention first of all. We then generate their interest. We create desire to speak to us. And then we call to action. We give them the offer of that call so that they say yes to. And then what we do is we make sure we have a leverage library. This is key. And so a leverage library is the bank of case studies that we've got, the testimonials, the results, and proof points across the business. Maybe it's awards that we've generated. What we want to do is AI will match each lead to the most relevant one automatically. You saw at the start of this video one of the case studies that I imported we took a brand from 700k to 4 million. And we're understanding what niche they're in. And so AI then understands when it cuz it lead scores and scrapes data on each lead, which case studies, which pieces of leverage to show in which email to the relevant leads. So, the email feels hyper-relevant to that individual. This will dramatically increase your response rate, your meeting book rate. Now, the data that we will scrape in order to aid AI with this process will be the company. So, maybe we've got the LinkedIn page, it'll be the revenue level if we can, if there's any public information if they're big businesses, the tech stack that they're using on the back end, any recent activity on social media or in the news. And then this allows for deep personalization, not just hi first name, I saw this company and thought of this. No, we're actually creating not just personalized first lines, but personalized emails following a specific framework. Now, you might be wondering, Jordan, why don't you just show me this email? Now, why haven't you shown me the exact waterfall? It should be pretty obvious. I'm sure you understand. If I publish the exact lead waterfall, the exact email that I'm writing for a real system, which is this, that is running right now, which our revenue is dependent on, then I would just simply be giving away all of my intellectual property and anybody that's got an agency that remotely competes with me will just copy everything that I have done word for word. And then it won't work anymore. What's the point in that? I don't want to stop being able to provide amazing value to you guys. And so I want to teach you how you can do it yourself without you just copying and ripping all of my directly. I think that's fair enough, guys. Um — I'm saying that as if you're you're saying why don't you give me your email? Of course, you understand that. You're not a fool watching this. That's why you're watching this channel. The quality of what the AI writes comes down to what you feed it. If you give it nothing, you get generic drivel. the leads company data, a relevant case study from your leverage library, and then the proper a copywriting framework to follow, you're going to get something that genuinely reads like a human sat down and wrote it for one person. That's the difference. The context is everything. The context is always everything with AI. And then every email gets checked by an agent before it's allowed to send, okay? So, every single email gets automatically written, but it also then gets fact-checked by an agent before it gets sent out. And so word count, spam triggers, how many questions you're asking. And if it fails any of those checks, they get binned and the system writes a new one. So, nothing goes out unchecked. But the bit I really love about this is the feedback loop. Most people think they can just blast a ton of emails and be done with it. But outreach is the same as ads. We've got to be continually testing. And so every 2 weeks, the system itself looks at its own data. What is getting replies? What is getting ignored? What's actually booking meetings with the right prospects? And then off the back of that, it generates new variations and runs fresh split tests. So, the winning copy replaces the losing copy automatically. So, without me touching it, the emails are getting better and better with time. All right. So, up until now, we've only talked about email. But email is actually just one channel in this system. Let's jump on to surround sound.

Go through the tiers. Now, all tiers get a free email sequence. They get the initial cold open. They get a follow-up, which is just a bump. And then they get a breakup. So, it's like hey, like we I've been trying to reach you. I can't get a hold of you. Before I reach out to some of your competitors, I just want to give you one last chance. It's a little hack there for you. That works really well. And we actually do like to do another little bump. So, we test another bump here as well. So, three follow-ups is generally speaking what we will test depending on volume. It depends on how well the second and the third follow-up are actually performing. And the first and the second follow performing, we will add a third one in there. Every lead gets this regardless of the score. So, it doesn't matter what the quality of the leads is. They would all get an email. The leads should only be quite a high quality if they're on the lead list anyway, but some people do leak in. Then priority leads with a score above 50, okay, will get the full email sequence plus they get a WhatsApp or SMS depending on their region and they get a voicemail dropped straight through to their inbox. These are the leads that most are most likely to convert and so they get more touch points or they're more likely to convert cuz the best ones are actually the top leads and so they get the voicemail and they get a WhatsApp or an SMS depending where they're from. The top lead 75% plus get everything plus they get a handwritten letter. And there are many different services that you can use online to do this. This is such a hack because physical mail in a digital world stands out considerably. In a world that's where it's easy to send emails, DMs, where it's easy to create ads and hide behind the computer, the stuff in the physical world is craved and it works incredibly well. And so the top leads are getting this. Of course this is a more expensive thing to add into that waterfall. But when we do this, when we have a multi-channel approach, we do this because most people only use email or DMs. They have these one channel that's competing with hundreds of other competing businesses. And so stacking channels compounds the response rate and the lead score dictates the investment per lead. All right, last bit. What actually happens when somebody replies? Well, we use AI appointment setting. This is the reply flow. When somebody is interested an AI will send a booking link, but it'll also suggest some times to that individual based on the calendar integrated with our calendar and it'll then book the meeting. That's pretty simple. But what about they say not now or not interested? The agent will objection handle that lead because it's got all of the context of our business, who we are, what we sell, who we serve, all of the micro nuances and will be persistent for as long as it needs to be responding to as many emails as it needs to in order to convert that not now not interested lead into either a booked call or just removing them from the system temporarily, but it'll be far more persistent than a VA will at massive volume of emails. The reason why this works is because of speed to lead, okay? If we get a reply at 2:00 a. m., AI will respond in seconds. A VA will respond in 8 hours and by then the lead is cold. This is especially vital when it comes to email. In fact, 20 times more likely we are to convert a lead if we respond within 10 minutes. But when we're sending 3,000 emails a day including follow-ups, we're getting a ton of responses. And so it's very easy for VAs, even a multi-team of VAs, to be very lazy with the way they objection handle and the way they respond when it comes to email. And so AI appointment setting with email, arguably more than anything else, probably the most effective place that you can use it because of the nature of how humans treat responses to email outreach. We've got to get that speed to lead in and we've got to come in with max effort every single time. Now, even if leads don't convert or they haven't engaged after the full sequence, we do have a 90-day cool-off period so we put leads back into the sequence every 90 days. They forget about us, but we don't just throw them through the same sequence. They re-enter with fresh messaging, a new angle, and a new case study. And so the AI will recognize, okay, this is the second or the third or the fourth cycle and so now we're able to hit them from a different angle. And previously we weren't able to do this or it was too high effort in order to track this. It was it was just wasn't worth the time when we were building manually, but now with AI it's much easier for us to bake in these edges when it comes to the strategy. So that's everything. I think that is the full system from start to finish. And look, I know this is a lot in this video. There's a lot of moving pieces. I'm not going to sit here and pretend that you can build all of this in a day. It took me a little while. lot of thought and experience in the space. But the architecture that I've run you through is sound and the principles work at any scale. Even if you're just starting off with a basic lead list, you haven't got many on there, you've got a couple of scraping tools, you've got a decent AI prompt, and a copywriting framework, then you are miles ahead of most people who are still doing this manually. And if you want help building stuff like this, then I'll put a link in the description to book a call with our team. I have been doing outreach for over 10 years. I used to dream about the moment that I could automate the whole process properly. And now we're finally here. It's actually quite surreal, really. What a time to be alive. Now remember if you want access to the Notion doc from this video, then follow me on Instagram at Jordan Platten. DM me AI outreach and I'll send you this entire document. You can use it as context as when you're building this system as well. And then also go and check out the next video if you want to continue learning. Subscribe for more and I'll see you soon. Cheers.