AI SEO: How I 100x My Business with AI (Full Guide)

In this 35‑minute presentation, Eric Siu, CEO of Single Grain, explains how the rapid drop in AI costs is reshaping marketing and SEO. He shows how small, AI‑augmented teams can build massive businesses by creating “sub‑agents” – specialized AI workers built with tools like Cursor, Claude Code, and Model Context Protocols (MCPs). Siu walks through practical workflows: using voice dictation (Super Whisper) to command AI to analyze Google Analytics/Search Console, generate SEO‑focused content, and publish drafts to WordPress. He argues that the future of marketing will look like a multi‑instance video game where each AI‑agent handles a specific function (SEO, CRO, finance, etc.). He introduces the concept of “Search Everywhere Optimization” (AEO/GEO/LLMO), urging marketers to diversify across Google, YouTube, Amazon, Reddit, podcasts, and short‑form platforms. Siu shares networking tactics (AI‑dinners, lunches) that build defensible personal brands, and he details how his team adopts AI through coaches, fanatics, and converts. He also covers free “vibe‑coding” tools (Replit, Lovable, ClickFlow), AI‑coaching bots (Lindy), and the importance of scaling quality content while monitoring token costs. The talk ends with a Q&A covering tool selection, content scaling, privacy, and ROI measurement, reinforcing that AI amplifies growth for those willing to adopt it.
Chapters
Please welcome CEO Single Grain, Eric Siu. So I now stand between you guys and lunch. I'll make it worth your while. So here's how we're going to do this. I'm going to go for about probably 35, 40 minutes or so, and I think we have 2 mics up here. I'm going to leave some time for Q&A as well, but there's a lot to cover. I'm probably going to overload you with a lot of information. We'll talk about the practical use cases around AI and really how you can 10x yourself and 100x yourself. Let's get into it. So just quick background on myself. My name's Eric Siu and main business is an ad agency called Single Grain and I have a LinkedIn ad personalization software called Karat. And then I mess around with my YouTube channel called Leveling Up. And then yesterday, if you were here, I was doing my little podcast with another marketer named Neil Patel and our podcast is called Marketing School. But enough about me, just some other housekeeping items too. That little orange box that you see, that is if you want these slides. So hopefully it gets good and you want to download them. You'll see that orange box continue to pop up throughout the presentation. So you don't have to, I guess, take a picture of everything. So the first thing I wanted to cover here is this slide, which is insane to me. So not even just marketing, when you think about business as a game, it's changed quite a bit. Those of you that have heard of Cursor, you look at Cursor's monthly recurring revenue, $0 to $100 million in about 21 months with 20 people. Wow. And then you look at Midjourney, $0 to $200 million in 2 years with 10 people. Lovable, $0 to $10 million, took 'em about 2 months with 15 people or so. So I'm not gonna go down this entire list, but this was just a little while back. You look at it now, Cursor, $500 million run rate, Midjourney, $500 million with less than 100 people each. Lovable on track for $100 million, and this is in July and they have 45 people now. So they're all growing very quickly. So what does this say? This is telling us that you're able to build big businesses with smaller teams now, right? So we all see these efficiencies happening. Hey, Thomas. So the cost of intelligence is dropping. You look at this graph over here, you look at DeepSeek versus O3 versus O3 Pro. So this is DeepSeek, this is China's open model versus OpenAI's model. And if you look at the very right, just look at the yellow one, that's O3 Pro, $80, and you have $40 for O3. And then look at this dropdown. Dropped from $80 to $40. And then I think the other one dropped from $40 to $8 or so. So this is all just saying that the cost per token is dropping quite a bit when we're using these, when we're using OpenAI's APIs. The cost of intelligence is dropping, which means we're going to do a lot more for cheaper. That's what I'm really saying at the end of the day. So, and by the way, guys, I'm not technical at all. I'm going to show you how I go about running these marketing workflows and these business workflows really, and how I've built this business operating system. And I think Everyone here needs to take advantage of it. But again, the game has changed quite a bit. The cost of intelligence is dropping. But if we focus this back on SEO, which is the premise of the title, 25 Cent SEO, I think the first thing to call out that's worth calling out here is the fact that traffic, website traffic is coming down. So those of you that aren't too much in the game of SEO right now, this is what's happening to SEO. So if you look at your Google Search Console, your website impressions are going up, but your website traffic is coming down, right? So website traffic is coming down. And we call this the alligator pattern. We call this the great decoupling. I like to look at this as, if you think about Uncle Warren Buffett, he likes to look at businesses that are coming down as cigar butt businesses, right? So I'm not saying that Google is dying. Google's actually doing very well, $115 billion in net income in the last 12 months or so, number one of all companies. But website traffic itself, that game is changing, but that doesn't mean you should throw it away, right? And AI overviews, this study was done by Ahrefs. This actually shows that AI overviews are affecting website clicks negatively by about 35% or so. So I wanted to set the tone with these stats in the beginning to show you how the game is changing and how we, I think the only thing now is for sure that we know is that the game is changing, but we have to adapt, right? And so when we think about my company, the people in my company, there's 3 types of people right now when it comes to the world of AI or marketing in general. You have A, the anchors, okay? So these are people that are resistant to AI. I don't want to deal with it. I don't believe in AI. It makes too many mistakes. It hallucinates. Yeah, it does all that, sure. And so you have the anchors, and then C, you have the converts. So the converts, these are the people that are open to learning. They just need some guidance, okay? And then F, you have the fanatics. So the fanatics, these are people that are constantly ahead of the curve. They're constantly testing new things, right? And what I would encourage you is It doesn't matter if you're a marketing manager, a marketing director, a CMO, it doesn't matter if you're a founder. I would encourage you, each and every one of you here, to set up dinners with other people that are at your level, similar to your level, maybe a few clicks above your level, and set up these dinners. And these go a long way. I'll just share a story here before I start to get into the stack. I think about 2 months ago, I hosted an AI dinner with one of my buddies, and we had a couple CTOs there, a couple CEOs there. That led to someone else inviting me to an AI lunch, and that had about 40 people or so. And that lunch, you have people doing, you know, all these different businesses, some are doing 8, 9, 10 figures or so, doesn't matter. What matters is everyone's a student right now. And when we asked in that room, who here actually uses Cursor? Even the non-technical people, they're using Cursor, right? And more on that in a second. But my point is, you don't necessarily need to have money to start throwing these lunches and dinners. If you can become the connector, everyone's going to remember you, one thing's going to lead to another, you're going to build your network, And nobody can ever take that network away from you. And that's how you're going to build defensibility with whatever that you decide to do in your career. And again, that brand that you build being that connector, that served me well. But we're not here to talk about networking, but just a little thing that you can do to help you. So this is the main thing that I wanted to talk about here. The way that we're doing marketing is changing. And I touched upon this a little bit yesterday in our little marketing school session, but I didn't have enough time. So I've got a lot of time today. So I'm going to try to go a little deeper and then tell you how I'm doing this. And then hopefully during Q&A, maybe I can, 'cause I talk really fast, maybe you guys can ask some questions, some clarifying questions, and maybe that'll help the entire audience here. But if you look at the very top, those 3 icons over here, who here actually uses Cursor? Just quick show of hands. Okay, so let's say it's like, I don't know, maybe 5 or 10% or so. And then Claude Code, anybody? Okay, wow, that's actually more. And then MCPs, anybody familiar with MCPs? Yeah. Okay, so a good chunk of people. So I'll explain MCPs real quick. MCPs are model context protocols. And just think of it as you can have your apps talk to each other. So you can have Cursor, which is your IDE or your coding editor, you could have it talk to your HubSpot instance, you could have it talk to your Gong instance, you could have it talk to your Google Analytics instance. When you combine Cursor, which is the coding editor, and by the way, I'm not technical at all, again, I can ask it to create sub-agents using Claude Code. So think of these as sub-employees, and then I can connect it with MCP so that the sub-agents, these employees, can call upon of different tools here. So I'm going to give you a couple of examples just to get your, your wheels turning here in terms of how you might apply this to your business. So one example might be I want to hook in, I want to create an SEO employee. So I'm going to create a sub-agent. I'll go into Claude Code and I actually dictate this using a tool called Super Whisper. So I'm not even typing this out. I'm just saying, hey, I want to create an SEO analyzer that will look at my Google Analytics and look at my Google Search Console and then also connect with the Ahrefs MCP, which is an SEO tool, and then tell me what content I should be going after based on what's converting on my website. So if I had to do that with my team, that would take what, days, weeks, sometimes maybe even a month or more to get that done. I no longer have to wait for that. I can get that analysis done very quickly. It'll say, hey, these are the highest converting sections for your site. These are the pillars you should be looking at. Here are trending keywords related to these pillars. And oh, by the way, here's content that's being cannibalized on your website. You should think about, you know, combining, deleting, you know, consolidating all these things, right? These are all SEO terms here, but My point here is that you can spin up these sub-agents, these little employees within minutes. I'm talking 3 minutes at a pop, okay? And then afterwards you can connect it with your CMS, you can connect it with your WordPress CMS and it will spin up a draft for you. And so doing these things, this was impossible for me a year ago and it's, to me it's extremely exciting. I'm probably spending a couple hours on this every single week and now what I'm doing is I'm actually for the fanatics on my team that are into this stuff, I'm hiring a coach for them to tutor them on this and to set them up with all this. I can go on and on about the different use cases here. I'll just give you one more and then I'll move on. You can tell I'm pretty excited about it, but connecting it with your HubSpot and then your Gong instance, show me the sales data from the last 6 months or so versus the last 6 months. What are the trends that you're seeing with both Gong and HubSpot? What should we be doing? Oh, you can connect it with all your other, your QuickBooks, you can connect it with whatever else that you have, but This to me is not just a marketing operating system, it's a business operating system. And the things that you can do here, and it's hard to see, but on the right, on the right side, you can see that it's giving me the keywords that based on keyword difficulty that we should be going after based on what's converting on our website. Think about how that works for you. And this doesn't just apply to SEO, this applies to all different disciplines. You can take a look at your competitors' ads. How are they running ads? What is their strategy exactly? What are the CRO tests that you should be running on your site to increase your conversion rates? All these things, you are now able to, you as one person, you're able to orchestrate all these sub-agents. So a lot of people like to talk about agents, but it seems like all pie in the sky. But this to me is we're getting a lot closer to it now. And anybody, I'm pretty sure most of us are not technical here, right? If I'm not technical and I can do this, you can do this too. And I'm going to show you how in a couple slides. So here's another example. This is just a simple example in terms of it. pumping out a blog post based on our writer guidelines. So it will scan our blog posts and it'll look at how we write our blog posts and it'll kick it out here. And then what will happen is we'll have people on our team, it's still very AI-assisted, so we have a human in the loop here. And when we have our team look at it, they're like, yeah, this is pretty damn good, right? And so I'm not saying you should go out there and use AI to create a bunch of slop, that doesn't work well, but it's already been proven that Google does rank AI content many times over and Ahrefs has a good, a couple good case studies on this. And so, This is the premise of this talk here. Like, I was just asking Claude Code here and Cursor, I was like, man, you're generating all these posts. Like, how much is this costing right now per post? And it did an estimate. It's actually 13 cents for one of these. And it's going to get cheaper and cheaper. You guys saw the token, the drops earlier, right? That's just going to continue to get cheaper and cheaper. And so I just look at all of this as if you and I can do this, you become 10x, you become 100x. And the way I like to break this down is, I look at money and I think money doesn't change most people. I think money amplifies your nature. I think AI amplifies your mind. So if you're not a person who wants to grow, which is everyone here wants to grow, if you want to grow, it's going to amplify you 10, 100x down the road. If you don't want to grow, I think it's going to make you 10% of who you are, right? And that's not what we want to do. So I actually think the future of marketing, the future of business in general is going to be multi-instance. So I used to play online poker in my college years, and I used to play, I used to have 16 tables open at once. This guy over here, he's a pro player, his name's Elky, and he would have 48 tables open at once. But just imagine it this way, each little instance that you have, one's an SEO employee, one is an analytics employee, one is a CRO employee, one is a finance employee. You, all these employees that you have working with you, you, and then if it tells you something's wrong over here, oh, we need you to review this over here. Okay, boom, you're going to open up that instance, right? So I think what's going to happen is, The future of marketing, the future of business is going to look like a video game. And to me, it's becoming really fun. Like, I think everything's a video game and this is exactly how I wanna play it, right? And so, you know, this is interesting. I was watching, there was an Instagram reel today with, it was on Diary of a CEO, Stephen Bartlett, and this investor, Mohnish Pabrai, who's known as the Indian Warren Buffett. He was like, Why do you need free time? You know, running a business, it's like having an orgasm every hour. And I thought that was a little cheeky, but to me it's like, if you're playing a video game, I don't know if it's orgasm every hour, but like playing a video game, this is pretty fun, man. So how am I doing this right now? All you need to do, literally this is what I did. I was talking to a friend about this and he's like, oh yeah, I just hired a coach, a cursor coach. I was like, why do you need a cursor coach? He's like, well, I'm not technical, so it teaches me how I can use it from a creative standpoint, from a design standpoint. So I was like, huh. How do I use this for my life? So I went on to Upwork, literally Upwork. That's all you need to do. Just go to Upwork. And I put up a— put out a job posting for this and I hired a guy from China. And this guy from China cost me like $50 an hour. And what was he doing? He was literally getting on Zoom calls with me and I was telling him what I wanted to do. And the first Zoom call was a little messy, is like a little introduction, but Zoom calls 2 through 4, All we did was he would do screen shares with me. Sometimes I'd share access to my control over my computer and he'd set things up for me. And by the 3rd or 4th session, the 3rd or 4th hour, I was already humming on this and I had all these different MCPs set up. And so it's really not that bad. If you're semi-technical, you can just set this up yourself, but give this 3, 4 hours and you'll become— you've got to become like a multiple of yourself at the very least, right? Maybe 2x, 3x, eventually 10x. eventually 100x. That is the premise of this talk. I have a couple other things to show you as well as we go through this, but I think if you're to leave this talk with anything, just understand the future of marketing is going to look like a video game. It's going to become multi-instance, and I think that's very exciting. Okay, so we talked about just how SEO is changing, how business is changing, a little more on how SEO is changing. So people here like to call it AEO, they like to call it GEO, they like to call it LLMO. Neil and I, we call it search everywhere optimization. And so 15 years ago, you might be able to say the only game in town, organic game in town was Google. There was no Instagram. YouTube was still just very much getting started. Snap wasn't there. Meta wasn't really there, right? So, but nowadays YouTube, second largest search engine in the world. Amazon, if you're trying to optimize for shopping, you got to do Amazon. App Store optimization if you're doing mobile. LLMs, you got to optimize for the LLM surfaces too. Communities like Reddit, for example, these, the graph that you're seeing move moving up to the right, these are all communities moving up into the right because Google is favoring those more. And so that means that besides the things I just showed you, we all need to become, even though I'm pretty sure most of us here aren't SEOs, my background's in SEO, but as marketers, we need to become more generalists. And that's why I'm even calling this out right now, the search everywhere optimization thing, because it can't just be one thing. Even though for traditional SEO, yes, you can still write transactional content, bottom of the funnel content that converts well, that still does really well for us. that can't just be the only thing that you do. You're going to have to diversify. And so I spent a lot of my time on nerding out on YouTube. And so this is a little interview that I did with Seth Godin a while back, about a year ago or so. And YouTube to me, because I think YouTube is top 3 cited by the LLM services, I think Reddit's number 1, Wikipedia's number 2, and then YouTube's number 3. And then depending on what you're looking at, sometimes LinkedIn even pops up as number 4. And so to spend time on YouTube is going to be a nice investment for the next decade or so. And going back to the MCPs that I showed you, the Cursor stack, the Claude code stack, this is what I'm doing with YouTube. I'm like, okay, I want you to look at my competitors over here. I want you to look at the last 90 days. What are their top performers so I can look at their packaging? Because about YouTube, for YouTube, it's about the idea. It's about the packaging, okay? And so it will, you can't see the output here, but when you download the slides, you should be able to see it. But it gives me the output on what's performing well. It gives me a little table here. Hey, here's what worked out well for Gary Vee. Maybe you should consider this format over here based on the last 30 days of videos that you did. And so now I have a strategist. Sure, we hire a lot of professionals from YouTube jobs like ytjobs.co for our YouTube channel. But sometimes I just don't want to wait. And guess what? I'll give you a little another nugget here. By using Google Nano Banana, I no longer have to wait for thumbnails. We have a couple thumbnail designers, but Sometimes I just don't want to wait. Sometimes you just want to publish something if it's newsworthy. So that's important. So you can use the stack that I showed you to help you as your YouTube strategist, all the other things that I mentioned. But just keep in mind, when you're playing on these other channels, whether it's YouTube, whether it's Reddit, I think the problem with SEO is, and I'm really kind of criticizing my space in the SEO world, is people, whether you're a paid media manager or an SEO person, you have to look at each channel with a different lens. You have to post natively for each channel. A lot of SEOs are like, oh yeah, you know, YouTube's really about, optimizing for the right keyword and, you know, the right tags and all that stuff. That doesn't matter nearly as much as the idea or the packaging or the first 30-second retention or your, the open loops that you have throughout the videos, right? And so just keep in mind, every single channel you look at, look at it from a native lens. Don't look at it from, you know, the industry that you come from. And that's how you're gonna win on these channels. And I think I was talking to a guy yesterday after our session and he's like, man, you know, you and Neil, you guys talk so much about the AI stuff. I feel like I have nothing to talk about. But I'm like, no, you know way more than us about product marketing You know way more than us as a head of AI. Why don't you start a head of AI podcast? And so I think anybody here can start creating content and it's just going to compound over time. And I actually was talking to a friend and we just, we're like, man, whatever we've seen usually with business or whether it's business or creating an audience, it always seems to take 3 years. And then someone else kind of chimed into the conversation. I was like, why do you think it takes 3 years? And my immediate response is most people just don't want to sit through the misery and not being able to see progress for 3 years. But honestly, that for whatever reason, every single business I've started or every single channel I've started, it's always 3 years. And if you can sit through that, you're going to be fine. And so I guess that's another takeaway for you. So we talked about YouTube. This is interesting to me because when you look at the new game of medium now, especially with VO3 from Google, which is their video model, and Google Nano, Nano, which allows you to have character consistency, which is a thing AIs have struggled with. But you can do this now. And if you have the rights from like a celebrity or a micro influencer, you can use something like Midjourney to, or not, sorry, not Midjourney, but HeyGen, I guess even Midjourney as well to make these ads for you. And this ad in particular from Cody Schneider actually tweeted this, is getting like 69-cent leads for a coffee shop. And so we're gonna see a lot more of this come into play because pretty soon we're not gonna be able to tell what is AI versus what isn't. And this example here, this is actually a naked grandpa skiing. So this is from the CEO of Mudwater. So he used VO3 and it looks like he used ElevenLabs for a Ryan Reynolds type voice, I think, or Burt Reynolds it says. And he scripted it using Notion. And this cost him $100 to make. It looked pretty good. He had some B-roll from his drink, Mudwater, and then he had a bunch of just, older people doing things that they want to do in their life, right? Whether it's starting a cult or skiing naked or something like that. And he said that typically this would've cost him like, I don't know, 6 months, 6 months of time, but maybe a couple hundred grand to get this done. So again, costs of everything are coming down. It's a matter of how much are we willing to play with this stuff? And again, I'll say with these AI dinners, like we have another one happening in LA later this month, but we're just constantly seeing who's on top of this stuff. And it really comes from the top. Like if you have that power, if you're the founder or you're an executive that can influence decisions, it's really going to come from you, that the excitement's got to come from you. Otherwise, it's not going to happen. This is also interesting to me too. So this company is based in SF, and they've been getting a lot of press, maybe not so much in the last couple months or so, but this company is called Clue.ly, and their whole motto is cheat on everything. So it's like a heads-up display that shows you how to ace your interview or maybe like a job interview that you're doing, right? And so the top left one here, I remember watching that video that I think that video actually has 4 million views now, but it was like this girl and this guy were on like an online date, like a Zoom date. And the guy was trying to talk about football, but the girl had no idea about like who Neymar was and all these things. And the guy's like, who's your favorite player? And then the heads-up display was like, you should talk about Neymar because of X, Y, and Z. So she did that. And then everyone in the comments, like 4 million views on that video, what tool is that? What app is that? What app is that? So if you're in consumer SaaS, this works, right? But I think even in B2B SaaS, you know, playing the organic short-form game, for whatever reason, because we're, most of us here are in B2B, we're always 5 years behind of the D2C world. But I will tell you that a lot of my B2B SaaS founder friends, we are trying to explore what we can do in this world. So I think one of my friends, he does like weapon screenings for schools to just keep schools safer for kids. And one of the crazy things he wanted to do was just walk into like, stores in different states with a gun and see what happens to him, right? And I'm not saying you should do that, but he's trying to go a little like outside the box here. And so I don't think enough people look at organic short form because it doesn't convert well. What I'll say is sometimes when we do the Marketing School Podcast, I'll talk about discounts and I'll get like 7 to 8 million views. Like, oh, Shake Shack has a free chicken sandwich, 8 million views. Oh, this Loro Piana store over here is like massive 90% discount in Italy, like 4 million views, right? And can I attribute it to any conversions? No, but I can tell you when I go to business conferences, when a couple people walk up to me, I can tell you that's a touchpoint. And the way I look at marketing now is very much the rule of 7 was, oh, 7 touchpoints back in the day. I think it's like the rule of 21 now. People have to look at your stuff quite a bit. And if that's a touchpoint, I'll take it. I'll take it any day. So I think it's worth it to study companies like Clue.ly. I think there's another company called Jenni. I think it's J-E-N-N-I. Take a look at all these companies, look at what they're doing, and figure out how you can apply it to your company. So Again, I started with MCPs, but I'm just kind of going down the practical, here's what's working today in terms of what I've seen. And then I know this guy Rowan Chung here. So he's got an email list on AI called The Rundown. And for him, this Instagram, he built it in a week and got 100,000 followers. But more than anything, you can see some of these videos here. So the views, 2.6 million, 1.7 million, 234K or so. He used HeyGen for this. This is just going to get better over time. And he just covers the latest AI news and he has an avatar doing it and he uses ManyChat to collect the emails. And so if you want to grow an email list quickly, again, leveraging organic social, leveraging short form, that's a way to build your email list very quickly. And as we mentioned yesterday, I think email is what, like a 34x return on investment. So it's still the highest channel of all the channels today. So as I continue on here, let's talk about how you can, because we were talking about SEO earlier, how you can rank highly from an AI SEO and GEO standpoint. And so this is not to toot our own horn, but whenever you search typically on most LLM services, what's the best marketing podcast out there? Usually Marketing School is going to pop up number 1 or number 2. And I'm not necessarily going to say that Neil and I are the best or whatever, but I think if anything, we've probably been one of the most consistent in the last 9 years or so. I think Gary Vee's been very consistent. We've been consistent doing that podcast every single day. There's a new episode and that's why it pops up. We think about citations. Citations matter, right? So it's not just links anymore. It's how many reviews are you getting? How often are you publishing? How consistent are you? How consistent are you? How is the brand sentiment overall? And so Neil publishes a ton of content, I publish a lot, and it's like, oh, these 2 that publish a lot of content already are doing this marketing podcast. That's why we're consistent. And so I think that's one of the key things around ranking highly from a GEO standpoint today, or LLMO, whatever you want to call it. And But all of that ties back into, oh crap, these LLMs also look at Reddit. They also look at YouTube too. And so those of you that saw the session yesterday, apparently Neil runs a lot of Reddit stuff for his personal brand, but he wouldn't reveal it on stage. I'll get it out of him on the podcast. So make sure you guys subscribe. So I'm going to keep going here because I got a couple more slides. So these— yesterday what we talked about was these vibe coding tools. So Replit is amazing, Lovable is amazing. You can use all these to build what you want. And what we love doing with our business is we love giving away free tools. So this is one of our free tools here called ClickFlow. And, you know, one of our free offers is free, you know, X amount of words. And these are agentic workflows that we have for our team because these are the issues that they're struggling with. And so you don't necessarily have to spend engineering time on this, but can you go go to Replit and Replit is allowing you to do a lot more now in terms of deploying and just driving things to production. But if you can get 80 to 90% there and then hand it off to developers, then you're going to have something that you can give away. And I will say that a lot of our free tools, a lot of our calculators that we published in the last 3 months or so, they're ranking very highly and they're driving a lot of visits. So that strategy still works today. And if I were to give you another strategy, if you can create custom GPTs, because Google is actually ranking OpenAI's custom GPTs really highly right now. If you can do that, that'll be great. Even better than that, if you guys have some budget, go acquire some ChatGPT plugins because that's what one of my friends did in a certain niche. And he drives a multiple on revenue in terms of each month from the plugins that the plugin traffic that he gets. So you can go buy plugins or you can create custom ChatGPTs and just have them rank higher. Or you can go create different tools and just give them away on your website and you can vibe code these things. I think people like to make vibe coding seem like it's, you can just create anything you want. I don't think it's that easy right now, but it's pretty good. This is just a screenshot of Manas over here. And the reason I like Manas, these are 2 AI products that we use. We use Manas and we use Genspark. So Manas and Genspark, they have these super agents. And the beautiful piece about these super agents is that I can go to, 'cause Carrot is our LinkedIn account-based marketing software, right? I said, okay, I want you to create, you know, long-tail transactional keywords that are high-converting for Carrot. And using Manus's super agent, it found 50 keywords. I picked 20 and I said, okay, you're going to have to take a look at our writer guidelines to make sure that the blog posts are high quality. And it kicked out 21 blog posts in about 25 minutes or so. And you guys might be thinking, oh man, that's probably crappy blog posts, probably crappy crappy quality, but reality was maybe like 90% quality and then we had to have a human in the loop there, but that's still pretty damn good for whatever we pay for Manus a month. I think it's like, what, $50 or maybe $200 a month? And so just keep these in mind. I mean, it's, people might say, oh my God, you know, you don't want to just spam the internet. Yeah, you don't want to spam the internet, but if it's 90% there already, it's worth taking a look at. So Manus is great. Genspark is great. Love both of those. Earlier I talked about Replit. So Now everyone here has developer skills. The conversations I'm having behind the scenes with a lot of founders right now is that, man, everyone kind of needs to become a generalist, which is what I said earlier. So this is reinforcing the fact that everyone here, we're becoming pseudo developers, we're becoming pseudo salespeople, pseudo marketers, pseudo customer success people. We have a lot more power in our hands now, and I guess that's why it feels like a video game. This is Carrot over here. This is just another one of our tools. But I will say for Carrot specifically, What we did was we created something that wasn't available before. And so you can vibe code these things, right? We heard from one of our customers, you know, one of our challenges was that we can't create ads quickly enough. Let's say we want to target 100 companies on LinkedIn. It would take us weeks to create personalized campaigns for all of them. We're like, okay, we'll just make a tool. So we made Carrot and we weren't sure about it yet. That's the thing. So we're now on search everywhere, organic reach with LinkedIn. So I just took a screenshot of one of our features. I threw it up on LinkedIn and 1,300 comments. Now I think it's like 800 likes or so, but maybe 150,000 views on this. And probably I would say 800 leads for Carrot and maybe 150 to 200 of them were ultra qualified. And so we're going to get to a place, maybe not right now with Vibe Coding, but we are going to get to a place where you just listen to your customers, go to all your Gong calls, Otter, whatever it is that you're using, listen to what challenges, hey, What challenges do you have? Okay, that whole magic wand question. If you can solve anything with a magic wand, you can actually solve a lot of these things now. You can vibe code something up, and I have seen a lot of these vibe coded products actually making money now. We engineered this from the ground up, but that's my point of showing you Carrot here. It's because it came from a customer conversation. We built something, we reinforced what we built by putting it up on LinkedIn, and then LinkedIn proved it out. So, We talked about Manus here. One thing I do want to talk about, we talked about GenSpark, we talked about OpenAI, but one product that's also really good, that's, those of you that run a business here, or even if you're a leader, Lindy has been great for us in terms of coaching our team. And what I mean by that is what we do for our companies, we want to help people prioritize, right? Correctly. And so every single day, let's say I'm basically required to kind of report in on, hey, here's the highest leverage thing that I'm working on today. Here are the 3 big things I got done yesterday, okay? And that's kind of our standup bot that sits inside of Slack. Now what this coach does, Lindy, it creates these agentic workflows and then Lindy will go in there and it'll coach people based on what our goals are and how they can work on higher leverage opportunities. And that happens every single day. And we're gonna continue to refine this agentic workflow over time. But imagine having a coach in all your meetings. We all sit in probably way too many meetings. A lot of these meetings probably aren't run efficiently, but imagine if there's a coach giving feedback on every single meeting and then we can tweak and maybe some people don't need to sit in a meeting anymore. Maybe, maybe if we just tweak this meeting over here, maybe it's going to be 10 times more efficient. Who knows? But I'm sure we're all sick of wasting time in meetings. Everyone's talking about AI right now, but people are struggling to prove the business ROI of AI. Now, with Single Grain, which is my ad agency, that is not an issue because we try to stay at the forefront of AI. We leverage AI not only to make our team more efficient and faster, but to make your business more efficient. And the things that we do around marketing, it's still our bread and butter. But when we combine that with our AI prowess, that's what allows our clients to grow a lot faster. So if you're interested in learning more, if you're interested in acquiring customers faster, go to singlegrain.com. Again, singlegrain.com, and we'll see you on the other side. And as we get closer to wrapping here, all the documentation that you have, in Dropbox, in Box, even all the Looms that you made, these are all infinitely more useful now because we need to context engineer how we're working with these AIs. They need as much context as possible, and you have that within all the documentation. So all that useless documentation that, at least for us, that was collecting dust, it's now infinitely more useful. And as long as we keep it updated and we keep feeding it into these custom GPTs or these projects, we're gonna have a lot more of these, a lot more useful subagents or a lot more useful custom GPTs that we can use to do our work better And faster. So this is where I think we're heading. I think a human AI operator at the top, and then you're going to have agents running sub-agents, and sub-agents are focused on ideally one task. Maybe one's on SEO internal linking, maybe one's on creating an article, one's on updating content, consolidating content. But this is how it's going to look. One of you managing 100 of these or 1,000 of these, and you doing what you want to work on and not having to do rote boring work that makes you want to, I don't know, Not go to work. So I think this is the way, this is one way I like to look at this. So the orange line over here where it's moving up and to the right, that's where you stack different agentic workflows, you stack different agents, and you keep compounding over time. And these only work if you compound your own knowledge, okay? And the yellow is maybe if you're doing a little bit, maybe it's one single AI workflow. And then the flat line here is if you don't do anything at all. Arguably, that line should probably go negative. So You build your leverage today. And what I would say is the bad news is that marketing is getting harder. I think it's very much a red ocean right now. But the good news is that you have a golden opportunity right in front of you and most people aren't going to do it, but you're not most people. So take advantage of this opportunity. And if you want the slides up over there in the, I guess, bottom right orange. And I guess we can take questions now if there are any questions. So you mentioned a lot of technologies. If you had only 3 and no other option, just had 3 to pick, who would be your top 3 that you cannot live without? Top 3 that I can't live without right now? Yeah. Probably that stack, the Cursor, Cloud Code, MCP stack. What are your thoughts on scaled content and whether or not it's gonna be penalized by Google, other search engines, ChatGPTs? I mean, it's, it's, everyone's using it. It's spammy. What are your thoughts there? Yeah, I don't think anyone has an issue with quality. As long as we're scaling quality over time, there's not a big issue with it. And Google has even said this themselves. And then Ahrefs did a study, I think it was on over 600,000 pieces and they compared it AI versus non-AI and it was virtually the same thing. And so that was my concern in the very beginning. And just like you, I'm sure I don't want to put slop out on the internet. So any original research that you can have is gonna do a lot better than just copying other people. So right here. Yeah. Um, so I run a services company. Uh, I am technical, so I've got the Claude code and MCPs all set up, but the idea of like transferring that to my team, given sort of the complexity of it, seems really like daunting and scary. So I was wondering if you could go into a little bit more detail about like lessons learned with that, how that worked out. How much you had to support them, like all of that. Yeah. So I'm sure you can see with your organization, like who the fanatics are, correct? So I target the fanatics first and then I just kind of float it to them because I'm constantly showing them what I'm doing. And part of the reason why I make this stuff on YouTube is to throw it out there and say, hey, what do you guys think about this? Right? And the ones who, when their eyes are lighting up, I'll say, hey, what do you think if I got you a coach for this and I paid for the coach, would you dedicate some time to this? And some of my best people right now, in the evenings, they're working with this coach. And I just use the same coach that I hired originally. And what I'm going to do eventually is I'm going to have it become like cohorts. But right now it's more one-on-one. So I think there's a crawl, walk, run to all this. Does that answer your question? Cool. Right here. How do you make time for content and get leaders to also make that time? Leaders to make time for content? Yeah. Talking specifically about video, podcasts, Those types of things. Yeah. So I, so I probably spend around, I don't know, 8 to 12 hours a week on content. Um, maybe a little less, but there's a day that's dedicated to content and then the rest is dedicated to operating. But the thing for, for myself is I, I'm working most of the time, like my mind's always working, so I, it's hard for me to break out, but I would say there's one dedicated day to content. Yeah. I saw your process for using Manus to crank out a whole bunch of blog posts and we hear a lot about using AI to crank out tons of hyper-personalized content as part of the loop marketing. Is there anything that you need to do or that you've learned to help get all of that content visible to AI? Like what's the submitting a robots.txt file for AI? Is there something that you need to do to make sure that AI sees all that? Yeah, people like to talk about llm.txt, but there hasn't been any evidence that it does much right now. I think there's literally, Ahrefs has been publishing a lot of great case studies on this. So as of right now, not really. It's just to me, and companies don't want to hear this, but when companies come to us, they're looking for GEO, AEO, all these things, a lot of it's still very much traditional SEO and it's not that different yet. It'll probably change in the next 12 months or so. Right on. Yeah. Hi, Eric. You mentioned we went from 8 to 21 touchpoints. Do you have any experience with that? Is something happened recently? And then the second part is, are you using any intent data to improve conversion and results? Yeah, so the first part is, so I just want to make sure I get the question right. You're asking again, the first question? You said we went from 8 touchpoints to 21, I believe. Yeah, so the rule of 7 used to be that you need to see something 7 times before taking an action, and now it's much more, I'm saying 21 because attention's so disparate, and when you think about the time spent on YouTube is actually the highest time on video. And then you, I think LinkedIn's number 2 with 26 seconds or so. But we're all doing different things. We're all so much more distractible. And that's why my, I haven't seen, I would say yes, I have seen evidence because it's 99% of people aren't ready to buy, right? And so you just need to be in front of people more, which is why people are creating content more. Yeah. And around intent data, yes. The short answer is yes. Yeah. You mentioned an MCP workflow with WordPress and having it write drafts for you from Cursor. And I use Claude Code. Can you speak a little bit more about that? Yeah, so with Cursor, when you create a sub-agent using Claude Code, it will ask you what is it that you're trying to do, those of you that aren't familiar with Claude Code. And so for me, I just use a dictating tool called Super Whisper, and I'll speak to it for about 4 to 5 minutes or so saying, hey, here's exactly what I want to do. I want you to use these MCPs over here. I want you to connect with WordPress, but first I want you to analyze my Google Analytics and my Google Search Console to see what's performing and what isn't performing. And then maybe in this case, I want you to create a blog post for me based on our writer guidelines over here. So I just dictate out exactly what I'm looking for and then it'll pop up the prompt and I'll look at it and I'll edit the prompt and then I'll try running it a couple of times and we'll just continue to refine it. Well, and more so the MCP side of it from WordPress, is that a particular plugin you're using? Yeah, there's, there's an MCP site that has all these MCPs that you can just hook in with directly. It's pretty easy. I just forgot what it's called, but it's pretty easy to use. Okay, I'll look for that. Yeah. Thanks. Yep. Hello, I use Semrush to evaluate our AI metrics, like share of voice. Semrush showed that we're ranking number 1 for AI Google search compared to our competitors, but we're in last place for ChatGPT. Do you know like the different factors that influence each LLM? Is that public information? Is there anything we can reference to to specifically rank higher for a specific LLM? Yeah. So right now there's not enough information. There's all these tools out there like Profound and they help you, like they show how you rank and things like that. But what I would say is, again, it's not what you want to hear, but a lot of LLMO and GEO stuff, the reason why we rank, it's still a lot of traditional SEO. It is citation, it is links, brand sentiment, but this is why you have to take the search everywhere approach where you are hitting Reddit, you are hitting YouTube and Maybe you nail down one first and then you scale it out to the other ones. I think if I were starting from scratch today, I'd probably lead off with podcasts and then do long-form YouTube and then cut it up in short form and then go try to make it, try to hit the other channels like a lot of people that you've seen. I'd probably do that because I know LLMs are looking at 30 to 60 different sources whenever you're searching for something. And when you do deep research, it's way deeper than that. So still very much traditional SEO. It's not what you want to hear, but it is time to play the search everywhere game. Okay. Thank you. Cool. All right. One more. Um, what, one really specific question. You say show up everywhere and Amazon was on your list. Do you mean sell on, like, we, we're in e-commerce. We don't sell on Amazon. Are there other ways to show up there that I should know about? Yeah. So Amazon's Amazon SEO versus like Pinterest SEO versus Google SEO, very different. Amazon SEO is a little easier. Pinterest is a little easier. But I just, what I, my whole premise of saying that was that you have to think about optimizing for the channels where your customers are. It's just marketing 101. So nothing game-changing there. Okay. One last question. When you're talking, I'm at the very early learning stage. So when you're talking about the MCPs and bringing all this information in, I can hear people in our company worrying about having that private information uploaded somewhere. Do you ever worry about that, or how can I respond to that? Yeah, you can't— I mean, you can imagine running it, uh, so you can, you can run this locally as well. Um, you can use models, you know, some people are like, oh, like, can you trust a Chinese model? I'm Chinese, by the way, but like DeepSeek, right? So DeepSeek is a great model that you can use. You can run it locally. Um, so a lot of these things you can do locally without having to worry about, you know, sharing your information out there in public. Thank you. All right, everyone, thank you so much for this.










