Codex Booked Me 4 Sales Meetings This Week - Here's How

Marketing agents are absolutely overpowered right now, and I'm going to show you the ones that I use to grow a lot faster. So we are talking about ones that will help when it comes to reviving deals, getting new deals, getting the right talent over time, stuff that can optimize your trial reels. There's a retention optimizer. There's a whole bunch of these different agents that you can use. By the way, I'm going to show you how you can get them if you stay towards the end. So let's get into it. So first up, I have our Single brain warm pilot outreach. Okay. So what this does exactly is it looks for deals that we might've lost before, or people that have come in as leads for some of our other products or services. And it just reaches out to these people. Right. And it also makes lookalike audiences off of these people too. So what it does, and by the way, this booked 4 meetings just last week by itself. So there's a lot to work off of here. And you can see I'm using ChatGPT here. In fact, I'm using Codex. And you can see these are the people that we're talking about reaching out to here, the different angles that it has. It checks if we have a reciprocal relationship or not. And it figures out the angle there. Right. And also you can see up here, I like, I like using Codex because you can put in your different browser. You can log into your browsers here. You can, you see I'm logged into HubSpot here. I've logged into email here so I can give it control of what I wanted to get control of. And then I can exit out of those afterwards too. And so if you look at this overall, you can see what it's doing. So like, here's what it's doing. And it's also recommending actions too. Like, here's what's happening. 5:30 AM, 6:30 AM, source closed lost, no call discovery, caret reciprocal Gmail calendar and lookalike candidates. So What these 2 really just mean over here is no-call discovery means somebody that we haven't talked to before that actually requested a discovery call, but we never just connected with them. People that closed lost. And so deals that we've lost before, maybe deals that we've pitched for the different products and services that we have. Carrot is just one of our products. And then reciprocal Gmail calendar and then lookalike candidates. So lookalike candidates might pull from ICP fit. So ideal client profile fit customers, and then it'll find lookalikes off of that, right? And then I can just help iterate off of it. And at 6:30 AM, it says produce one 25-row packet with role size, warmth, workflow fit, safety, email And copy, right? And so then I can choose to approve or disapprove. Now, I like this because I don't want the agent to just YOLO and kind of do whatever it wants. To me, an agent is still very much you tell it what to do and it does the thinking in the middle, and then you have a human in the loop to review, right? That to me is, is kind of the, the go-between right now where I don't think it's completely safe to just let it do its own thing because there can be reputational damage there, right? And so as much as people like to say there's this panacea right now with agents, not quite right now. But that to me is what an agent is. It's not just, not just like an automation that you're running, right? Like a Zapier or any automation. Like, that's not it. You got to have something that, that's of thinking in the middle and then also making decisions for you and executing judgment there, right? Another benefit I'll give you here is that we've tied this into our Loops Creator skill that we have, and I'll show you a little bit in terms of where you can get all these things that I'm talking about. But the Loops Creator skill will make this into a loop that will continue to iterate on all these agents that I'm going to show you, right? So this is a good starting point, and you can see over time it's just going to give you these packets. I can dictate out what I like, what I don't like about it. You know, if it's reaching out to the right people, the wrong people, and I can tweak it. And you can see I could just dictate to it over here. Another call I'll give you is I do use the /goal command to build these things out. And then right now, as of right now of this recording, the meta right now is to use 5.6 Luna Max instead of using SOL 5.6 Medium Plus for everything. Luna 5.6 Max is really good for this stuff. So that's what I would say about the first one. And that is already a really good starting point if You're thinking about not just recovering revenue, but growing your revenues. Now, the next one is a GEO AEO workflow that you can use. So the way I think about this is when you are creating content, I create a lot of content around my podcast, my YouTube channel. Internally, I'm having a lot of conversations with the team, whether it's internal meetings or whether it's customer conversations or helping with sales calls. Those are all conversations that I'm having. And in there, there's a lot of spikes. There's a lot of philosophies. Now, what's to say that we shouldn't be ingesting that content to figure out what type of content we should be creating from an AEO standpoint, from an SEO standpoint to our site, right? Because we want to drive more awareness. And oftentimes the stuff I'm talking about is newer stuff, right? We're talking about, you know, building a company brain, we're talking about revenue-focused company brains, there's all these things that should be created into top of the funnel, middle of the funnel, bottom of funnel type content. And so the way this works is this is a closed-loop content factory. And so it leverages a lot of different connectors that you might have. From Google Analytics, from Ahrefs, from Google Search Console. And then we use the ClickFlow API and MCP to crank out the content. And by the way, that's free for you to use if you want to use it. But this will then— this addresses the problems around content teams losing leverage in these 5 places over here. So why was an idea chosen? Rewriting the same ideas, not having ideas to begin with, right? And not having a measurement plan or like a publishing plan, and then not figuring out how to repurpose this stuff, right? And so what we've done with this one is I've actually handed off to my Hermes harness to continue building, and then I've actually handed off to engineers, right? So oftentimes, A lot of these agents that you build, it might be good for yourself, but then you got to think about how can you actually scale it out to other people, not just on your team, but for the broader public to use. Because this is now something that we want to build into the ClickFlow product itself. And so this is just something that you should be, you should be thinking about because there, there, maybe you're not creating podcasts or YouTube content or whatever, but for sure you have customer calls, right? For sure you have sales calls. For sure you have internal calls where you can pull from Gong or you can pull from Granola, for example. They both have APIs, they both have MCPs that you can use. That's how you actually build more leverage, not just from an AEO SEO standpoint, but content across the board. So you can actually daisy chain a lot of these things together, and then you can, you can decide to run them on recurring, and then boom, voilà, you have an agent going, right? That's how you want to think about this. And, and one thing I'll also say too is, you know, the word agent sounds scary, but to me, if you can combine these, if you combine these different skills that you have These different workflows that you have, and you invoke them at a given time. And if you combine them together, okay, great. Well, you're now having an agent solve an outcome for you where you're asking it to do something, it's thinking for you in the middle, and then a human is reviewing at the end. That's all this really is at the end of the day, right? And then if you want to get it to a loops level, well, that's for another conversation, right? If you want to build loops around it, that will improve over time. That is the AEO SEO content operating system. Now the next one is you got to build an outreach command center. So you can do outreach from a sales standpoint, from a marketing standpoint, from a recruiting standpoint. So as we look at the outreach command center here, here's what it's doing every single day. All right. Single brain. It's, it's doing that sourcing that I was just showing you. Okay. So when I have a command center, it's easy for me to look at this from a high level. TalentOS replacement pool. So looking for the roles that we're hiring for right now, AEO, SEO, AI automation engineer, sales engineer. Or sales leader, you know, 4 deployed AI operators. These are all things that we're looking for. And then also we are, you know, we want this to record into our applicant tracking system and CMS. And again, I'm using Codex here because it can, I can log into these different things and it's, it's, you can see it's actually using LinkedIn Talent Manager up here. And then it can use my email if it needs to use it. Right. So I've built enough trust with this, or I trust this enough now where I can just let it run. You're going to have to continue to tweak it over time. And by the way, this has been running for a bit now. It's not like I asked this to run. You can actually see it's running right now. So you can see here's where the blockers are, whatever it is exactly. But my directive for this thing is just move the ball forward every single day. And I'm very specific with who I'm trying to reach out to, whether it's from a recruiting standpoint, who do we want? We want people that have 2 promotions at 2 different companies with an average tenure of 3 years. So that's very specific on who we're looking for because that's actually a good history, a good barometer of performance. And then I can click on these little proof documents over here, but I just I tend to want the packets to be ready for me to review and then let it go. And then what happens over time is that this is a loop that runs every single day and it gets better and better each day and it tells me the tuning that it makes. And so that's how you actually improve on this stuff. And so to me, this is outreach. We're talking about this from like a marketing, from a business standpoint. These things are all important. We're doing outreach all the time, right? We need good talent, we need good marketing, we need good sales, right? We got to bring leads in, we got to close the leads, right? We gotta, we gotta bring in more amazing people to fulfill our customers too. And that's what it ultimately is. Everyone needs to create a trial reel optimizer agent because it's going to help you get a lot more organic views. And so as an example, if we're looking through this right now, you can see that we have ABC. Okay. We have your P&L may be hiding. These are the different overlays that you can see these hook overlays over here, right? Your P&L may be hiding $500K, same numbers, different decision. And your costs look fine, check this, right? And so what this is doing is it's actually connected to my Metricool API, which is connected to my Instagram, and I'll look at trial reels that it should be running. So look at what's not having good retention or videos that we published recently that could have better retention. And then it will set these new overlays and set them as trial reels, because that's a test that's worth running, right, over time. So you're trying to get more mileage from the videos that you're pumping out. And the cool thing here, like you can see current controls is 38%, 3-second view rate. So how do we get that to You know, higher, 50, 60, 70%. 70% is like you're going viral. But at least maybe we can target 50, 60%, right? But the whole idea here is that we originally, I simply just connected, I saw a couple of Instagram videos on Callaway, who I love following, right? And I just put his stuff in here. I was like, hey, we need to optimize for retention rates here. How do we do that? How do we set that as an optimizer that can run every single day? This is important. I'm showing you this right now because this is, this is real. Like, let's, let's, if we click on the video, let's just see how it goes. My AI CFO made us $500 grand on its first Stay till the end and we'll actually share a skill. Okay, so not bad, right? I didn't tell it to optimize much. I didn't really give it a lot of instruction. So let's just take a look at this one too. My AI CMO saved us $500 grand on its first month. Stay till the end. Okay, so it just, it really just added the overlay and that was it. But the cool thing too about this is when you're using whatever harness you're using, whether it's Codex or Claude Code, they're both great. And I pay for both of them. When you, when you think about this, it's like, oh, when I was playing tennis the other day, I was using my Meta glasses to work on my crappy tennis game. I said I had 68 videos or so. I dumped them all into Google Drive and I said, hey, Codex, take a look at them and find the best rallies. Okay. Where— and it found maybe 5 videos that were actually really good where we're going back and forth when I was playing with my friend. And it edited it perfect for Reels and I just posted it to Reels and that was it, right? So there's so much that you can do with this now in terms of the powers that you need. And don't get me wrong, I still use my Hermes harness all the time and I'll make more videos on that. But this Codex by itself right now is so powerful. Same thing with Claude Code. It's so powerful right now. with the stuff that you can do with it. So I think it's nice to have a retention optimizer and a trial reel tester. Okay, that's what you ultimately want. And then ultimately, you're going to drive a lot more views. And so I've been paying more attention to this now because this is a game that I feel like I haven't exactly cracked. And I want to crack the game and I want to, I want to get my team to focus their attention on this. So a trial reel optimizer, it's good stuff. Now the next one you want is a thumbnail agent. Okay, so this thumbnail agent, check this out. So I just put in a video here, give me 3 more videos based on this video. And I think this is a video that I made already. And so by the way, This is me wanting to optimize the videos we had. And when you work off of this one, you could actually create a side agent off of this one. Okay, so let me give you a sense of how this works. So it worked for 14 minutes and the packaging was low quality, so I asked it to continue to try again. So it gave me, it gave me these, and actually these were the ones that I said were— these look pretty AI, like bad quality overall. And I said, these are not good, like try harder, right? And so I said, these look a bit too AI and they don't quite make sense. Okay. And so it pumps these out over here, which in my opinion are still a little bit too AI. And I said, but now it's not quite aligning with the brand of the Leveling Up channel. So keep in mind, it's these skills and these agents that you put together, you're going to have to continue to refine them over time, which is why I like the concept of loop engineering. And so now this looks better. This looks more on brand ultimately. Am I going to say they're perfect? No, they're not perfect, right? But you got to iterate on these things over time. And then from there you can decide, okay, You know, can I say, uh, use the Loops Creator skill and then make this into a daily routine that you're running where you're optimizing the videos? Absolutely, you can do that. And then what you can do is you can just tie— you can log into your YouTube Studio here. You can just open like a, like, like a tab here, and then you can just log in and, and, and do what you need to do, and then just have ChatGPT working there, right? Or Codex working there. So to me, these are all high-leverage things that used to cost a lot, that like, if you wanted someone to optimize or run trial reels for every single reel Not practical. If you wanted someone to optimize every single one of your videos each time, if it starts to tank, not practical. But now it's practical, right? You just need to have someone on it. And what I would encourage you to do to all these things that I'm showing you right now, you got to make sure that once you get it 50, 60% there, if you have the benefit of having a team around you, or you have resources around you, hand it off to those people and see if they can take it, right? And just see how far they can go with it. Because they're— if they're ultimately responsible for driving views higher, then they should be paying attention to this. Now, by the way, if you're enjoying this right now, all the things I'm talking about, if you want to make these agents, you got to go to the Skills Dojo. Okay, so skillsdojo.com. That's where a lot of my skills are. And then you can take those skills and you can refine them, and then you can make them into recurring routines and you can combine some of these skills together. And then you actually have an agent that's working for you. You can use our Loops Creator skill in there to make loops, and then they can improve, they can tune every single day. And that's what you ultimately want. Check it out, skillsdojo.com. Okay, so this next one that I like calling out is this Mr. Siu Sponsorships Queue. Okay, the reason I like this one is because you can see over here, it's basically made a little CRM for me. Okay, and then also it just every single day, it's like, hey, because I get about 20 to 30 inquiries a week on sponsorships, I don't want anybody on my team dealing with it. I don't want to deal with it. So someone on my team doesn't have to deal with it, right? In fact, I can just give them this. And so you can see all the sponsorships coming in, and then basically it will draft emails to send to all these people, draft replies. And it'll get better over time too, right? And so what happens is they'll say, oh, you know, if we review this draft over here, okay, the standard— hi there— the standard sponsorship placement with Mr. Siu is $25,000 for one X, LinkedIn, or YouTube community post, including 30 days of platform-native paid amplification on the selected platform only. It does not include editing, sublicensing, cross-platform perpetual usage, exclusivity, or custom production rights. Guess what? All this stuff over here, I never added that. Through the Loops Creator skill that we have, it added that. Again, this is, this is working already, but imagine if you hand this to someone on your team, maybe it's an assistant, maybe they can just deal with this over time, right? And I don't even know what this, this 10 minutes expected, but maybe this is like negotiation back and forth, but maybe this is like, you know, money saved and time saved or whatever it is exactly. All these things that, that, that were just not practical before, they're practical now. And when you think about something like this, okay, we're talking about revenue generation from AI. Okay, well, what if I can get one of these per month? Or once one of these per month would be $300 grand a year, one every 2 months would be $150 grand a year. That would cover a lot of my costs on the content side. That's how I'm thinking about this as a business owner. But then you have the command center for outreach. Like just today, I spoke to some amazing people that were— that came from the outreach system that we have. Okay. And I told you, we booked 4 meetings from the single brain thing that I showed you earlier. Okay. We're running all these tests right now from optimizing thumbnails, from optimizing Trial reels, like retention, 3-second retention rates and things like that. These are all things that lead to more awareness, to more leads, to more customers that we have to fulfill. And that's the virtuous cycle that we're playing here. But the key thing I would say to you is you got to make sure once you get this to a certain level that you're handing this off to people on your team. In fact, the reason why I'm making this video too is for people so I can send this video to some people on my team and say, hey, are you ready to take this on? Can you take this on? Right? And then you want— that's why you also need to hire people who are very AI-forward too. that can take on this stuff, right? And the name of the game here is that you keep iterating, you keep tweaking here. And by the way, let's say you hand it off to someone. Maybe when you see this becomes a bottleneck again, maybe in 6 months you come back to it, you help them iterate on it a little more, and then you go back to something else. So I hope you enjoyed this video. Let me know what you think, and we'll catch you in the next one.