Why The Forward-Deployed Marketer Is the Most Desirable Marketer

So, your job as a 1% marketer is you're constantly looking at the bottlenecks, right? And then you're improving your bots because let's say you're deep in one area such as paid media, you would know all the things to do to make the best paid media bots out there, okay? So, your job is, let's say you're specialized in one area, go make the best bots out there and then release them to your team. And then maybe other team members can then participate and then you all can make the bots better because I truly believe, again, that we're all just supposed to be bot managers. We want to hire more for deployed marketers. And 4-deployed marketers for us are people that work on their— they work on marketing. They're deep in one area, right? You might be deep on SEO or paid media or copywriting or creative, whatever it is exactly. But then you understand the disciplines at a high level, right? So you're broad and kind of at the T and then the I of the T, that's where you are deep, okay? Now, a 4-deployed marketer also works directly with the customers too, so they can figure out Based on whatever solution they build for the customer, what can they take from that and put into the scaffolding that they have, the foundation that they have for their workflows, right? So we are actively hiring more for deployed marketers. I'm going to show you my Slack channel in terms of how I'm trying to work with it from an AEO SEO standpoint and how I'm already starting to try to hand off my bots. Okay. So some of you that have been watching these webinars for the last few weeks or so, you have seen that I have been Making these bots. I've been tweaking these bots. But if you're doing that and you're just managing all that by yourself, you're not going to be able to drive true leverage. You need other people on your team to be able to do so because you might be able to get it to 70%, 80% there, but it's truly the people that are on your team that are in the trenches. They are going to have a lot more context than you have and they're going to be able to tweak it to make it even better. Okay? So let me share my screen over here. So this is our Slack channel, right? Okay. This is our Slack channel, and then you can see in this Slack channel we're just testing this right now, right? So we have we have these different agents inside this this this channel. So Doodoo over here is my ChatGPT dot. Okay, so it has a lot of context from how I build things, and it uses Astra. So when I needed to do heavier thinking inside of Slack, I have it do that. Okay, now Team Geo Audit. What does this do exactly? So this finds the SEO Geo citation gaps, and it writes fixed briefs, and that so we can basically attack competitor gaps when it comes to A. AEO, or GEO, whatever you want to call it. Then we have the Team AEO SEO bot, which I'll cover in a second. And then we have the Team SEO AEO measurement. Okay? So I had these sitting privately as bots inside of my GrokBot, but now I've introduced them inside of Slack. So we are now, even though GrokBot graciously hooked up all these accounts, now we're just paying for them all, right? It's so good to the fact that our team— there's like an uproar yesterday when GrokBot, basically they had some issues with it, right? So that means that they have product market fit here because people are really mad when When it's not around, right? So what I would say is this, when you look at this channel, what it's really doing here is, uh, from an AEO SEO standpoint, is it's looking at content that I'm putting out from an external content standpoint or internal content standpoint, and then it's mapping out what we should perhaps be drafting to the blog from an AEO SEO standpoint. Okay? So a headline here might be OpenAI Dots versus Grokbot versus Muse, which agent should you buy? Or B2B content formats ranked S to D tier, why case studies beat short form for leads. Okay, and how to stop AI ads from burning $100K before anyone checks, or lead sources ranked S to D tier for B2B referrals and emails to win. So I, I'm familiar with all these because I covered these in various short-form pieces of content, or even like my, my podcast on Marketing School, or my leveling up videos. So what happened over here is Jev— Jev is a classifier, it's an API that you can use where it can basically classify things very quickly for cheaper. It can classify things very quickly as well. It basically looked at our evaluation for, hey, what should be considered clear? So it suggested 12 ideas initially and Jeff only passed 7 based on the criteria that we have. Okay. And then you can see over here, it's like, oh, hey, we actually— here's the tape where we were talking about this, right? Could be from Granola Notes, could be from YouTube, could be from YouTube over here. And then it also shows the closest tape, right? Basically meaning, have we published this piece of content already? So my point in showing you this is that this is nice, but it's too much for you to manage, especially if you're a marketing leader or if you run the company, you can manage this. Sure, you can create all these bots, but if you have to be the one that's spinning all these plates all the time, you cannot work on the high-leverage things that only you can do, right? Maybe you're getting a deal done. Maybe you're recruiting someone. Those are only things that you can do. If someone else is not doing their job ultimately, well, then you let them go, right? So, well, if the bot isn't doing its job, then you let it go, right? But if the bot is doing its job and you have someone that you can hand it off to, well, then you promote those people, right? That's ultimately what you want. So me getting it in here has been unlocked now because I talked to the team this morning. I said, hey, I kind of set the marching orders for this and you can see a lot's going on in here and we have to continue to tweak the channel, right? So Haley in here is like, hey, what do you want to have by the end of the week? And do you want published content or mainly polished content? She's like, I would like to prioritize this. Isaac's already working with this in different ways now. Until today, we haven't had this, or actually yesterday, we hadn't unlocked this. Now, let me show you a point of value over here where I got my value. Okay? So it's like, okay, hey, these posts over here look like they're good to publish. Okay. Programmatic SEO and AI Search Visibility Playbook for B2B SaaS. That's not bad. Jev versus your SDR queue. How we cut 60,000 leads to 2,000 ICP fits before outreach. Okay. That's not bad. So these titles are getting better. But the problem is if I'm spinning like 20 different plates at the same time, I can only allocate so much focus, right? But if I have one area of focus, I can power that up a lot more. But if I have like 5 or 10 or 15 areas of focus where I have to keep managing different areas, I can't get things done. That's why in your company right now, if people can't get things done, then you have to do their job ultimately, right? Which is not a good thing. And even if they don't know how to do their job and you hand them a bot like this, they still won't know how to do their job, right? So you still probably have to let them go, which you don't want to do. So this is why I say the top 1% of marketers, they work like this. They build leverage, they build bots, they hand off the bots, and then they move on to other areas. 99% of people are not going to build bots as of today's recording, okay? Because keep in mind, it hasn't even been a year yet. So what did I do in this case? Look, there's 50 replies in this thread. So I said, hey, you know what? I was like, okay, okay. Um, I was like, Haley, can you, can you review these over here? Give me your thoughts. Okay. I was like, hey, does this one look pretty good? And you know what she says? Hey, to be honest, this looks very AI. It doesn't look that good, right? And I was like, okay, well, that, that's good feedback because I'm not going to have the same level of detail or care that she has when it comes to, let's say, editing a piece of content. All right, so, so then it's like, I'm like, okay, hey, can you take that feedback and tell me what you would actually intake? So see the link above. So here you don't see me getting offended. Here I see this as being able to collaborate with my team on the spot allows us to make idea babies, and these idea babies are much better than just one, one person's idea. I'm not going to pretend that I have all the answers. Okay, so— And a 1%er marketer doesn't have all the answers. So she says all these things over here and, uh, she's like, okay, you know, based on the feedback that she gave, here's what I would actually lock from an eval standpoint. So that's the definition of success. So she said— so what she said was like, hey, you know, the title must pay a promise, you know, stats must already be named, they must be hyperlinked, you know, we need to have an anti-robot voice cap. And then she asked me today on a call, she's like, are you willing to let it talk in like first person? I'm like, yeah, I'm fine with that. She's like, can we do this? Can we do this? Can we do this? I'm like, yeah, we should do it, right? Because I'm not that close to the work. And so what happened over here, I was like, you know what, let's have it compete. I said, hey, show me the before and after HTMLs of these files, okay? And I'll show you guys in a second. And tell me what we would codify from this set. And so, okay, it's like, okay, Team SEO AEO bot starts working again and it's like, okay, well, the before version is a 58 out of 100, which is an F. Now, the after version is an 86 out of 100. Hailey's comments impact are 84 out of 100. Is systemic for this set: undefined tools, fake anonymized reviews, title promise drift, uh, tip tile— tip pile mismatch, robotic voice, weak Yoast score. All right, so let's take a look at this. I've just— this is my first time looking at this, by the way. Okay, so this is the before version. Let's take a look at this. Uh, you know what, it's not designed well, so that makes it harder to look at. So Let's see what this looks like. Okay, we're just going to look at the intro here. We're just going to look at the text. Okay, so here's the before text. Okay, so before text, quick overview. This playbook is for growth and creative ops who want AI video ads without turning every render into a hero project. The job is running Sea Dance videos into Meta ads as a manager agent job. Clear inputs, clear outputs, review gates, and a kill switch. The mechanism is a pipeline that generates variants packages. The AI, right? I'll always say it's AI. You could just read that, right? Now let's look at this one over here. So quick overview. On the 11,000 Ads in 1 Month episode, Cody Schneider and Eric Siu walked through an AI marketer workflow that can push creative volume into the thousands. The punchline is not raw volume, it is— Okay, this still kind of reads like AI, but what's happening now is the team is getting in here and they're like, oh hey, we actually have some anti-robot, anti-slop skills over here that we can use. And I think we could make this a lot better. So she gave her feedback, that was just one round, but your team has probably put together skills or they have other things inside of a GitHub repo that they can insert into this bot. and make it even better. I think this opener is a little better and it looks like it starts with the embed over here and then it just goes more. I would trust because our team already works with clients on this stuff that they're going to do a better job than I will. That's how this looks like Grok bot inside of your Slack and your team can interact with it. The other thing that I've done here too is I said, hey, you know what? I don't want Isaac to be able to talk to this bot yet. I want him to sit and watch for a little bit. Haley, I'm going to give her the managerial access where she can make changes to this because if you move this a channel and everyone can make changes to it immediately, it kind of just becomes like this Frankenstein thing, right? So you probably just want to have one person on it initially and then you can start to roll it out to other people inside of the channel. And that's how you expand this stuff, right? And by the way, you want to continue to clean this up too because I can't read all this stuff. I'm just like, what is all this stuff, right? And so I think I had said to it like, hey, I need you to give me tables where I can actually read this stuff. Otherwise, I don't understand what I need to approve or what I need to disapprove, right? And so Your team can not only help you fix the functionality with these bots, but they can also help you continue to upgrade the user experience inside of this bot. So this is what it is. Initially, look, I couldn't read this over here. I was like, hey, man, I can't read this. Can you make it into a table? I still can't read this table over here. And then I'm like, can you just make it into an MD file? Boom, now I can read this. And then I'm like, okay, we need to continue to tweak the reporting. So Again, if you're a top 1% marketer, you know how to tweak these things, you know how to play with these things. And the way my team looks at it is like the team that's hanging out in here kind of messing around with this right now is that we're basically patching this all the time. Because that's like, guys, like we have ClickFlow, I get it, but like this should be ClickFlow. ClickFlow is our SEO AEO product, right? It helps like there's a lot that it helps with from like a tracking standpoint, from a share of voice standpoint, like you can see your decaying content. That's all helpful, but the way I see it now is that we're all kind of builders so we can keep patching what we're doing. We can keep patching our software all the time. And it's not a job. That's just for software engineers. So, I'll leave it at that and then I'll check out— Jay says, true, but if you spend too much time on X, you might spend all your time chasing hype and not focus on output. Well, yeah, but you should be spending at least a couple hours a week learning and then implementing things that actually move the needle. So, Jay actually brings up a good point here. So, a lot of people, because you can build anything, they start to build everything, right? And then they act like they're doing something with it. And so a top 1% marketer knows that whatever it is that they're building relates to their top priority and they are driving revenue or they're driving— let's say they're driving pipeline, they're driving revenue, or they're driving more traffic, right? Ideally, it's something that's more tied to revenue or pipeline. So if you can do that, you behave more like a top 1% marketer because a bottom 50% marketer is going to constantly be saying, well, look, I did this. I estimated that I saved this much of this amount of time over here. My estimated money savings is this over here. And they just start to say all this stuff and you start to get confused by what they're saying. And they start to try to show off the slop that they build. A top 1% marketer is always thinking about, how do I make sure that what I'm building shows my good judgment and shows that I've actually thought this through and shows that whatever I'm trying to present, I'm not just presenting the AI-generated piece of content. I'm not just throwing a slop grenade over to the next person that I haven't reviewed, that I actually can't I can't back completely with all the words that let's say a Claude Design has put together, right? So, top 1% marketers, that's how they behave, right? They have that bias to action. They know that throwing a slop grenade is rude and that's what it is. And by the way, a top 1% marketer understands these things. How do you read a cohort chart? How do you read a P&L, right? How do you read— how do you understand what How do you understand how to optimize for growth rates, right? Or net revenue retention, I should say, right? How do you optimize for usage? How do you look at where the funnel drop-offs are, right? And then you can ask the bots to make those reports for you so you're getting them. And a top 1% marketer, they know how to cull, meaning that they know how to cut things, right? So for all these new bots that you build, you're going to be cutting a lot more, I believe, in the very near future than you are creating because a lot of stuff that you make could become outdated very quickly. And, and by the way, like, these models are always leapfrogging each other, right? These harnesses are always, are always kind of leapfrogging each other. But I, I do believe right now, um, I think GrokBot is amazing. I think if you're a top 1% marketer using something like a GrokBot— don't get me wrong, I, I still love Hermes, I just don't use it that much anymore. It's probably more of like my strategic brain, like my individual brain. Um, I'm gonna show you, um, how I've been messing with Muse as well because we had a meeting earlier this morning with, with Muse. I will probably show you a little bit around. So we got Grokbot, we covered Muse. And those of you that haven't been watching this, I'll show you some of the other critical bots that I have. So Grokbot, Muse, Instinct is interesting, but that's more of like a personal thing. And then I think, okay, I feel like I'm missing something. There's Grokbot, there's Hermes. Oh, Dot, right? Dot is amazing. So let's maybe Let's maybe pop over to Dot for a second. When I manage Dot, let me give you a good example over here. I just go to town on this bot and I basically just use it as my primary chief of staff bot. And then there's a little call button inside of ChatGPT where you can just keep talking to it. And I would say that the voice mode for this thing has gotten better and better. I also like just inserting it into random channels. I— we have one channel where it shows like all the people that are joining on our, on our email list, right? And obviously you can just, you know, set like some type of enrichment on your own and just connect this with their like an email API key. But all that to say is it will show me, um, all the people that are opting into our email list that are ICP fit or M&A fit, because we're looking at potentially acquiring more agencies now, probably around 15 to 80 people or so. Um, that's kind of the range that we're looking in. Um, And then also, so we talked about M&A, we talked about ICP fit, and also recruiting as well. And then you can use something like a Jev, or now OpenAI just came out with Decisions API, which is open beta, and it can classify these things very quickly for very cheap. And so to me, being able to insert this into any channel I want and also collaborate with my Grok bots, and also we have an open channel where people can ask any questions, I can insert this in there. And sometimes when I see people ask a question, I can have this add more context around it, but it only responds to me, so it's secure for me. I think OpenAI is getting to a point now where it can have your dots, right? So this dot over here can collaborate with other people's dots on the team and you're going to be able to make multiple ones. So I think this agentic game is going to just continue to explode. So you have to continually remind yourself as a top 1% marketer of your number one priority, right? And so, because if you don't, you're going to try to do everything in the world, right? So what is your number one priority all the time? What's that one KPI you're being held accountable to? And even if your boss comes to you and they say, oh, you know, I want you to work on X, Y, and Z, you should just push back and say, well, I'm happy to work on that, but I'm focused on A, B, and C right now. Do you want me to rejigger my priorities? Because we only have so much time in the day. So this 2x2, this shows that people who do not have good judgment and do not use AI are dead weight. But people who do not have good judgment who use AI, these are slop cannons. And then also people who have good judgment who do not use AI are steady hands. But eventually if they keep doing this, They will become dead weight. Okay, right? So you're left in a scenario where it's like you need to use AI and you have to have good judgment. Guess what, Jev? If you help classify things at scale, if you can codify that judgment, will allow you to scale good judgment. Okay, in some cases. And then you have turbo brains, right? Use AI and good judgment. And so where does good judgment come from? Well, experience working, right? Good judgment also comes from using this stuff. Good judgment also comes from having good habits, having good learning habits, right? Having good health habits, which is a completely different topic, right? But if you have those things, then you should be able to continue to be a turbo brain. By the way, this X article that I'm showing you right now, I dictated this, right? And like this image, I'm like, oh, the image sucks. You use the single brain aperture design. Okay, boom, we have this, right? And so, I'm going to mess with this a little more. And I, I've like, I described my second pass, but, um, it's not bad. Like there's some images in here. Oh, by the way, like it'll chop up some, some of, uh, I didn't ask it to pull from this specific, uh, video and these moments from the video, but it found these videos, right? And so this one has 2,000 words. Now guess what? Ideally a top 1% marketer is not going to just publish this. Ideally the top 1% marketer will take a look at this and be like, okay, I'm going to edit this. Maybe this is 70% there, but this is where I apply my human magic, my taste, and my judgment to this to make it a top 1% post. Because right now, if you just post this right now, you're not a top 1% marketer. You are actually firing more slop into the universe because you thought this was good enough. You thought that just because you took a transcript from somebody else and you use an AI skill, that it's good enough. It is not good enough, right? All I'm saying, guys, is that whenever I spent more time, um, crafting the actual thumbnail or the creative itself or whether it's like a written post or not, it's done a lot better. Whenever I've just used an X long-form article skill and I just published it, it hasn't done that well. It's when I add my taste and judgment, especially for the first, you know, like the hook, right? And the first opening 30 seconds for a YouTube video. When I work on that, the video becomes a lot better. When I don't work on it, the video doesn't do that well, right? That element is really important for a top 1% marketer. And again, if you're not trying to be top 1%, well, you probably leave this webinar right now, right? But, um, top 1% marketer, they actually will put the tender loving care before and after, um, a piece goes out. All right, now the other thing I would say is, uh, with like a GrokBot, for example, okay, or any of these agents, you can't overload it with too much context, meaning that you can't have like 20 tabs of computer use running where it's controlling your computer, right? Because what happens is it's going to eat up memory on that computer, whether it's like a GrokBot computer or your personal computer, right? You can't do that. There's a couple other lessons over there. But guess what? You don't learn that stuff without testing. So you can't expect everything to do everything for you. So we talked about Muse a little bit, but I don't really have a lot to show you there. My only thing to— my comment to those of you that are maybe newer to watching this is that if you're using MetaMuse, try to use it on your organic bookmarks or organic social, and then it can do quick analysis there, especially if you plug in the Jev API there. And you can see all the influential people that follow you as well, and you can figure out who you want to partner up with from a sales standpoint. You can do that, right? You can also do it to connect it with your Meta ads, connect it with your Meta API, and have it spin up draft ads, uh, the ads for you. And, and, um, yeah, do that. So you can manage it, your Meta, from like an organic standpoint and a paid standpoint. And you can also figure out, uh, from a partnership standpoint too. And the thing I just gave you, like scanning all your followers to figure out who's influential and who you can partner up with, who's the Who should we recruit? Who should you do like an M&A deal with? You can do that for X. You can do that for LinkedIn. That's assuming that you have a followership there, right? Or you can even run it on your competitors too to see what's happening there, and you can reach out to those people. So there's that. Now, talking about Jev, what can you do with it now? All the people who engage with your competitors' ads, oh, guess what? Maybe you can reach out to them because you can classify all those people now and you can put them into like an outbound sequence and reach out to those people. All those people who engage with you on X or LinkedIn that seem to be ICP fit, or recruit fit or M&A fit. You can classify all those people now. The people who have engaged at least 3 times in the last 30 days, maybe they're higher priority, right? The people maybe 1 time, maybe not so much. But the whole idea here is that we can do a lot more, right? Which means our TAM in terms of what we can— the types of customers we can serve in our business, their businesses, they're getting bigger. But also what we can do, our capabilities as individual people, we can do more now. So Johnny says, you said you weren't using Hermes anymore and GrokBot. Why is that? My Hermes project manager agent can read/write to our team in iMessage, Slack, email. Then we have an SDR connected through iMessage who can do outbound. You think I should move to GrokBot even though the PM agent's working well and getting stronger over time? No, I think if it's working well for you, you should keep using it. I don't think you should just pull it out. My thing is Hermes has been a— it has all my context and it's a really good strategist. And also, by the way, when you use Hermes, you actually save a lot more on your OpenAI tokens or your Claude tokens, just an FYI. Because their harness optimizes for it. So just keep that in mind. So I still love Hermes and I use it in different areas. But I found that GrokBot, when it comes to the routines, when it comes to just getting things done from a reliability standpoint, has been solid. It's just been more— it's been easier to use and it's been more secure. Because here's what I'll say too, because you're paying You're paying SpaceX for this, right? They're going to pay a lot more attention to security and all this other stuff. And then when Hermes goes down or there's issues with Hermes, you usually have to troubleshoot that with— your team has to troubleshoot that for you. Grokbot, you just say, hey, support, can you just fix this? And so I think there's some value in that. So I still keep Hermes. It's good, but Grokbot has become number one for me and it's become number one for a lot of other people. And I love the fact that you can share templates and I love the fact that Teams is pretty helpful as well. Again, Hermes is great. GrokBot, I think, is just more reliable right now. Muse is good for its own purposes. I think ChatGPT, the dots, they have their own use cases as well. For me, it's just using more intellectual firepower with that harness and being able to voice chat with it and get things done that way. Anyway, I hope that helps. How are you managing context if you load all those skills? Do you silo them somehow? With the Grok bots, they seem to be able to take on a lot more. That being said, I wouldn't try to load too much context or too many skills onto one bot. So I would say Grok bot is the first time where I've probably made a lot more— just like I've separated responsibilities with the skills. And so for AEO measurement, I have one bot for that. For the content production piece, I have a bot for that. For talent outreach, We have a bot for that. We have for all inbound applications, we have a bot for that. For our agentic recruiter, we have a bot for that. And so I try to not throw too much on it because what I've learned from Hermes and OpenClaw over time is when you load it with too many things, it just starts to not work and then you have to repair it and you have to reset it. And that's probably the biggest pain in the butt. How do you train your team on AI and incentivize them to use it? AI, you said, can be a huge leverage, but in some cases can become a distraction. What I've learned from this, Samuel, is The top 1% of people already behave like the top 1% of people. You cannot train the top— someone who's like top 50% to become top 1%. You can't just magically train someone to want to work hard. You can't magically train someone who wants to become magically curious, right? And I've said this a couple times already but you have to find people who are innately curious already, who have already gone through that personal growth journey. And I do believe that's a personal growth thing. I believe that most people are most people. Yeah. because they haven't gone through that personal growth journey and they instead like to point fingers. So what I would say is we did, Samuel, to answer your question directly though, we do do hackathons, we do do AI office hours, we do have these daily standups now every day where we're actually focused specifically on these pods and just solving problems there. But also at the same time, if people aren't willing to meet us in the middle, we're not going to— if they can't meet us in the middle, then we're not going to extend that bridge because it's just not fair. To the organization. It's just not fair to other people. And what I would encourage you to do is after you finish watching this, go get the link from this and just say, hey, you know what? Take the transcript, right? Go to youtubetotranscript.com, okay? Or you can just plug this video in. Sometimes the harness will just watch the video. And just say, hey, Eric said a lot of things, but what's the one thing we should do from this webinar? What's just the one thing that's highest leverage as it relates to what we're trying to accomplish? Because your harness knows what your goals are. It has a memory as it relates to you. It has your context. And so, that's how I do a lot of these things right now, right? If I hear an interesting conversation, I'll try to screenshot. Maybe if it's like I'm reading something, maybe I'll just screenshot it. Or maybe if I heard like a snippet of a conversation, maybe I'll try to just send that snippet over to my harness and then we'll try to figure out how to Skillify it, right? But there's so many ideas out there today and I'm just like, Instead of me being on X and just bookmarking stuff, now we can just do stuff that seems to be relevant, right? But I would say more than ever now, prioritization is top one. Like a top 1% marketer is world-class at prioritization and they don't get distracted easily. All right, cool.