← Leveling Up with Eric Siu
Loading summary
Transcript1 lines
- [00:00]
A
Openclaw might be the closest thing to an infinite money glitch that I've ever seen. Not just personal life, but business. And not enough people are making enough content on how Open Claw can drive real revenue. Not just for yourself, but for your business. So for example, for us, I mean it has booked meetings with companies like Google. It has revived deals for us that were dead. So there are a lot of opportunities that I want to show you that you can use yourself. These use cases, you think about it for a second, you watch this video, I guarantee you stay till the very end, you're going to get some ideas from for yourself and it's going to help your business grow faster. Let's get into it. All right, so the first money making use case I want to show you is the deal resurrector. So the deal resurrector will go through my entire CRM. It'll look through what my sales team has worked on it. Look at all the gong calls as well. So that's our sales intelligence tool. And look at this. So with Champion right here, it says, hi Mike, hope you're doing well. It's been over a year. So it's identifies the time frame and when we, when we last spoke to this person. I remember the project got pushed because of the site restructure and some internal operating changes at Haynes. How's that for personalization? That is phenomenal personalization. Okay, here's how all that played out. By now you're probably on the other side of it. Since last spoke, we built some new capabilities that might be relevant. AI powered marketing agents that run SEO, paid and content 24 7. Not just tools. We've been generating 130 plus optimized pages and hours for clients versus the weeks it used to take. So I mean, I wouldn't say we have, you know, these are 24 7, you know, agents that we have running. No, I think that's kind of overselling things. So you might have to adjust this piece over here. But this is a, for the most part, I think this is 90% there. The, the, the bullet points just need to be customized. Right. But it is trying to pull from the direction in which we are moving into. And so that, that's why it's adding that in there. But it, for the most part it got it right. And so I can, if I choose to hit send email here, it actually will send through my email account. If I want to edit it, I can go ahead and edit it. And then not only that, if I want to find other people that work at this company. So for example, if this guy's email BO it will find those alternative people. Let me give you an example of this. So check this out. Good context from OpenClim. I'm telling you, this is your. If we're talking about infinite money glitch, the things that it can do right now, not only will it do the research for you, it'll do the work for you. Okay, so check this out. Haynes brand was actually acquired by Gildan in 2025. The restructure that Mike mentioned was likely related to the acquisition. Okay, so maybe he was restructured out. Who knows, right? So e commerce supervisor, relatively junior, he may have left during the acquisition. Who knows, right? And so then it shows me three better contacts to try over here. So you know, this person, creative and content director, most relevant for marketing services. This person over here for marketing, event manager, probably not the right one. And this person Leads champions entire U.S. collegiate and teamwear business. I don't know if these are the right ones, but at least it's moving in the right direction. If I prompt it more, it can actually do that, right? And so by the way, you might think, oh my God, this is so, you know, how do I go about creating this, right? All I will say is literally, if you take this part of the YouTube video and you say, I want to create something like this, you know, that's at least a good starting point for you to go with, right? Like when we're piloting, we're piloting this for us, for our single ring clients. Literally what we're doing here is we are building out these systems. We're testing them out first and we're making sure they work. And then we're compounding. We're basically building marketing teams for our clients that continuously learn with them, right? And then we have a four deployed marketer that helps with them, right? So if you're interested, by the way, in something like that, that is something we're piloting over at Single Grain. Okay? So use case number two is I have my open claw, which I chat with in my telegram, and I say, hey, what's, what's all the cool stuff that we worked on recently that would be cool for an X post? So an X article. So those of you that aren't aware yet, but X formerly known as Twitter articles are getting a lot of reach right now. So check this out. The article I put up yesterday, 56,000 views. The one before that, 70,000 views. The one before THAT, 103,000 views. 87,000 views. 80, 85,000 views. By the way, these were 95% AI done. I just assisted the AI with the hook. I handled some of the, maybe the opening lines just to get people hooked in. And that's all I did, right? Because my openclaw instance knows all the cool things that we're working on right now. And I just say, hey, what are some outside the box things that we're doing right now? And when you click into one of the posts, for example, how my OpenClaw creates expos that average 85,5000 views, what does it do? It shows you cool these diagrams down here as you go through it. It's structured well, you see, can it flows well as a blog post and it's not even that long. And look, 1200 bookmarks on it. Okay, so what did I do with this? I basically whenever I have something interesting and I say, hey, let's talk about how we're, we're averaging, you know, a good amount of views per article on X. Okay, great. I get an article that is cranked out and then I just, I schedule it here. I use Gemini Nano Banana Pro to make this, the, these there's little graphics so you can see we have these like Keish plush toys. And by the way, like I got the plushie idea from this one. And so the whole idea here, by the way, my original graphic that I had that only had 770 views was this one. But then I continue to use Nano Banana and then it morphed into this one. And then I wanted a first person view, or I guess this would be third person view. And then I wanted to add the open claw plush toy. And then the plush toy. I'm like, hey, like let's make it into an actual plush toy over here. Let's add a brain to it, right? So the whole idea here is that I'm basically learning in public and I'm showing you how I do this. And all I do is, you know, I basically, I continue to feed my open claw. I say, hey, here's how I talk, here's how I write. Look at my best X post over here. Here's how I write for my long form posts. I want you to be able to do that too. But by the way, I want you to add these diagrams as well. Okay? So those of you might be saying, oh, show me how to do all these things. At the end of the day, the best way of you learning this stuff is actually do it yourself. What you're really trying to glean from this is really the use cases, the ideas here and that's going to help you compound a lot faster. All right, so this brings me to my third use case over here. So if we look at my comm central, these are, this is me talking to my brand team and these are the ideas that my open claw gave me this morning. So I have a, there's a method that we use called the 4 on 4 content method. And basically we just look at, you know, content has been performing well in my space where the packaging has worked really well. You know, here's a good example. So idea number one, my AI agents made 47k last month. And by the way, this is, these are just ideas, right? These are not real numbers. Hermosi Systems videos get 2 to 3x channel views. How cloud code replaced. My entire team already performed well for you. Revenue proof videos dominate B2B YouTube right now. Effort is medium. Screen recordings on Oracle Flash Cyborg Revenue attribution thumbnail. So by the way, so look, it gives me a hook, it gives me a thumbnail. It shows me the impact and the effort. And there's already data. Like first it's looking at my data. It's also looking at Data that's on YouTube as well or this one over here. I fired my marketing team at Hired AI. 6 month update. Impact is very high. It's controversial and proof based. 80 to 150k view potential. And then here's the thumbnail that 6 months ago I released most of my marketing department with AI agents. Here's what actually happened to our revenue. So not a bad hook. Okay, so the whole idea here is I'm getting these, I'm getting these ideas. So for YouTube, me making this YouTube video right now, I was really thinking through the ideas first. It really is. The ideas are really the drivers at the end of the day. And you know, it's helpful to look at what's working well in your space. And so sure this will look at, you know, maybe what is happening with you know, Dan Martell videos which do well or Hermosi videos that do well. But it also looks at people in the AI space too. It looks at creators in the AI space. So it's not, it is based on kind of YouTube legendary YouTube creator lore or their, their teachings. But you know, it also looks at what's working well in my space as well. So that's why I'd say and look. So the intersection I'm looking for is really AI business and marketing because not enough people are talking about that, right? Now, and this is under exploited. Nobody's showing real business P and L impact from AI agents. Everyone demos, toys, nobody shows revenue. Right. And so the, the things I'm talking about before I move to the next one is we're talking about, okay, booking meetings with a Google, for example, okay, that's huge. Or reviving deals with like a Hanes or a, I guess it's called championware. Right. These are deals like our, our team is actually reviving these deals right now and those deals are moving through the pipeline. And, and bear in mind, I mean, this thing, I've only been using it for a couple weeks or so, it's only going to get a lot stronger. So the whole idea here is I'm trying to show you these use cases. I understand, I want you to understand that you're going to continue to compound with this thing over time. And that's why this is such a big opportunity that we're starting to pilot this for our, for our clients. Just a quick break. If you want to run LinkedIn personalized ads and landing pages, you can't do it on LinkedIn ads right now, but you can do it with carrot. It's www.care. that's K A R R O T.AI again K A R R O T.AI and we'll see you on the other side. Okay, so, so the fourth use case I'm going to give you is the business strategist. So this is, this is huge because openclaw actually has the ability. I mean, the contextual memory it has is really good. So it sticks with you over time. It knows what all your goals are and it continues to grow with you. Okay. And that's really important. And so what I would say is, you know, I am trying to brainstorm right now whether I want to buy a bunch of Mac studios for client service purposes. And so why do I want to do that? Well, I want to buy a bunch of these Mac studios because I want to run these. I want to, I want each client to, maybe they have their own Mac studio. Okay, so let's say it's an enterprise client that we're working with. And then they want this marketing team that grows continuously with them, which is like, for example, my agent squad is like the, you know, they're named after like, you know, Alfred from Batman, Oracle Flash, Cyborg. Right. These are kind of, you know, heroes from the D.C. world. So what I would do is, I would say, okay, well, help me brainstorm with me. And then also, by the way, if we're selling to an enterprise, there's going to be SOC 2 compliance in some cases. If we're selling to health, we're talking hipaa. You know, legal teams won't improve sending proprietary data to OpenAI Anthropic. So we might run these models locally on a couple devices, but maybe the clients have their own dedicated devices as well. So. Because they want their data kind of secured. Right. And so, you know, local models would be our moat. And it kind of breaks these things down. And then I'm kind of going back and forth with it, by the way. So it's saying, okay, you know, you can do 5 to 6x max studios, 30 and 36k. But I was like, wait a minute, I think it might make sense for us to, to lease the Max Studios. What do you think about that? And we go back and forth because, you know, total hardware, if we were to buy like, I don't know, 15 to 20 of these would be 100 grand or so. Okay, so you know, Apple Silicon demand is increasing. M5 Ultra will likely be priced 10 to 15% higher than M3. My recognition buy, don't over buy ahead. And then section 179 and everything. Pitch the security angle hard. Okay, so I'm like, wait, what? But you haven't considered a leasing angle. So if we look at this, okay, ROI is 10x in one year with three clients, I think we would honestly have a lot more than this. Okay, so then I'm like, okay, why not just lease? Because Apple has a leasing program and I can always trade it in for newer hardware in the next two to three years. What do you think about that? You're right. I was too quick to dismiss leasing for your situation. Lease. It actually makes sense because the whole idea here is if you're leasing a computer in, let's call it two to three years or so, you can actually trade it out for new technology or you can just choose to buy it out. So AI hardware is evolving fast. M5, M6 Ultra, M7 Ultra, every generation is 30, 50% more performance. Machine you buy today is outdated in 18 months. Leasing lets you ride the upgrade curve. Cash preservation. 12 grand up front on two machines is fine, but at phase two, six machines at 36 grand and phase three, 100 grand or so. Leasing keeps your cash deployed on growth, not hardware. Tax treatment is equivalent. So lease payments are 100% deductible. Okay, so it, my point is you now have a business strategist that's working with you and eventually it creates a memo for me on this and then I can show my team. Because, look, I've done the work with it, right? So you have an infinitely patient business strategist with you, and it actually can do the work if you want it to do the work for you. So if I were to say, hey, go out there and contact the people for me at Apple, go ahead and do that, right? Or even yesterday, I'll give you a little bonus here. Even yesterday, I was working with it to help me event plan. So I want to do a OpenClaw meetup in Los Angeles. So let's take a look at this. So check this out. When I work with my Alfred, which is my chief of staff, I was actually doing research on how I can throw a open claw event in la. And I wanted to do it in la. And I'm just like, okay, well, um, what's the speaker strategy? I'm like, can you help me plan? Can you write out the. Can you create the event page for me? Can you figure out, hey, what's our angle for this? You know, what's the structure going to be? Like, what are all the logistics? And it basically just became an event planner. And so in this case, it's like, okay, well, here's the speaker strategy you can use. You know, there's obviously, there's a lot of people in my network. Like, there's, you know, my YPO network, my EO network, and then there's a lot of entrepreneurs I know in la. So I can throw this just for entrepreneurs, because that's my, that, that, that's the approach that we're taking right now. And then I can also go to Open Call Discord. I can promote it on X and Twitter. And so, you know, my whole, the whole thing here is like, I'm starting to work with it. Okay? So instead of saying, I, I'm going to put this off to another time, I can able, I'm able to just do a lot more, you know, a lot faster. I was like, oh, I can speak. You know, my CTO could, could, could do this as well. And, and maybe we'll keep this small for the first one. Maybe 50 people or something like that. And so Clockcon is like, hey, how did it work for, for, for these other open Call meetups? How did they do it? So it's like Claucon was like this, you know, know, SF demo night was like this for your event, maybe. Let's just keep it simple. That's okay. I'm like, okay, go ahead and, you know, draft all the things for me, I'm like, let's, let's set this date. And I, I, I, I was like, oh, why, why don't we just try to be a little more, let's go a little more hardcore. Let's go Mediterranean food. It's gonna be a lot more expensive, right? But the whole idea here is I'm like, I'm like, okay, well, we don't do beer, but we do 5:30 to 8pm or so. The whole thing around this is that it basically it can plan the whole thing end to end, it can reach out for me and it can post for me end to end if I let it. But I would recommend having a human in the loop. So my point here is it's a strong business strategist. It could do event planning for you. I mean, really, sky's the limit. Don't let your mind be limited to just thinking about how it can help you from a personal standpoint. All right, so revenue use case number five. We got the agentic roundtable here. So these are all my agents that are reporting in in the morning. They also show a deal of the day and they're, they're basically talking to each other as well. So they're feeding each other context. And so earlier I talked about this, this Hanes deal over here. So it's the first thing that I called out. And you can see here, it's like, oh, okay, well this deal was worth 168k to start 14 months ago. We lost this one December 2024. We never followed up with them, by the way. $10 billion company and then context project delayed due to internal restructure. Perfect revival angle actually to send reconnection emails. So by the way, again, this can do all of it for you if you, if you let it. And then the other piece is this will also look for angles for, for content repurposing or maybe newsjacking. So Microsoft AI chief claims 18 months until all white collar work is automated Claude SaaS Apocalypse cowork plugins. So I can choose to attack these if I want, but I'm good with just showing you all how I'm working on this right now. And if you all enjoy this, like, let me know what other use cases you'd like to see because I can cover those. Okay. It also tells you conflicts where the system is broken too. And then if I say, hey, like, do I want you to fix it? I just hit the do it button here. That's it. That's it. Okay, so that's why this is crazy. And then earlier I kind of showed you the YouTube 4 on 4 ideas over here. I'm going to. By the way, look, I can choose to approve. And the key thing here is, like, it is important for you to, you know, indicate what looks good versus what doesn't. Because if you don't do that, then, you know, the, the. Your open call instance is not going to learn over time. You want it to get better over time. Now the last piece with this, you know, talking about the roundtable, there's Mission Control down here as well. So Mission Control is my dashboard that I use. And you can see that we'll call out a bunch of highlights that we should maybe attack. And then there's issues where things have, you know, not worked out. Right. Okay, so this is my instance of Mission Control where I can see, you know, the agents that we have working and the different models that they're using right now as well. And then the actions, the outputs. Right. The cost. And by the way, I mean, this is, this is great to me because I also have an activity log too, where I can see the things. By the way, I just added this yesterday and so I can see how we're. The amount of work that we're doing, the amount of work each agent is doing over time. So I can sort it by day, by week. Oh, come on. That's so cool. This is my first time seeing this, by the way. I can see kind of like a blog that we have this activity log where I can see the changes that it's making over time. And, you know, today I didn't get much time to work on it, but like every day I'm trying like 15 entries on the 15th. And, you know, the whole idea here is that you're just continually looking to compound this system. And the whole idea here is that hopefully there just becomes more and more over time. Right. So this is really exciting to me. And, you know, I wonder how far back I can go. Can I go to January? Let's see. Oh, that's so cool. So this is, this is analytics that actually works. That's really cool. I don't even know what this knowledge graph thing. Oh, I do know one. So I added this into literally, Sky's the limit. I said, hey, I want you to show me knowledge graph of all my different agents that I have. Okay, so Oracle for SEO and we have one for Flash for, for social, and then we have Cyborg for recruiting. And then Alfred over here. So you can see, you know, we have the different sources. The green ones are the sources where we're inputting data. And then there's different cron jobs, there's different agents and things like that. So you can see how everything is connected. Mission control is good because I have a Kanban board here to see what's active and what isn't. The different skills I have because I'm going to forget the different scripts that we're running as well. So this is going to naturally improve over time. And I just want to show you, like, I think everybody should have something like a mission control because it keeps you on task so you don't forget things. Because you could build anything, just not everything. All right, so use case number six is from a recruiting standpoint. And get this, we've already had amazing conversations with people sourced by this thing and we pay for a handful of other recruiting tools, other AI recruiting tools, and they're great. But the fact that this is working really well already is incredible to us. So for example, it will show us senior Director of Operations candidates and define kind of why they think these people are our top picks. And then maybe middle of the pack and maybe not so much down here. And then the cool thing is it's going to get better over time as we give it feedback, right? So we can react and give check marks, maybe X as well. And I can even respond to Cyborg directly within my Slack instance. So you saw me using Intelligram. I've actually just recently added it to Slack. So then now our team can start to work, you know, with, with Cyborg here on. From a recruiting standpoint. So you are building. Think of it this way. You're literally trying to replace anywhere from 1 to 5% of the chores that your team is doing every single week. And that's going to compound over time. And you saw the graph and I showed you a little earlier that, that those the agents should be doing more and more and more work over time. And, you know, that's really exciting. All right, so the seventh one I'm going to show you, last but not least is Oracle. So Oracle is amazing because it is our SEO bot. And check this out. I told it to come up with some outside the box ideas it can use to grow my traffic on single range. So from an SEO standpoint, from an AEO standpoint, and this is amazing to me because look, it identified 6,800 pages that are dead weight on my website. That's 84 pages that are 84% of pages that are dead weight clean. This lets Google Focus the 925 pages that actually work stop the negative quality signal. 1.84 million impressions with zero clicks tell Google's users don't want to use this site. Right. Facts are indexation of new pages. When we launched those 50 new, new agency pages, Google will pick them up faster on the clear, cleaner site. I don't know about that. Given that all the slop that's being produced on the Internet right now and this will take 60 to 90 days. Why it's not a 9 or 10. Okay. There's higher leverage things. I'm like, okay, well what items do we have to tackle on your list? And so we have these 50 agency page brief pages. We have internal linking architecture, we have an entity audit. We have 128 decaying page rescue plan 414 high impression zero click analysis schema markup specs, weekly tracking system. And so this, I mean this is what you would expect out of an SEO strategist. And again, it can do the work for you. And look, you can, oh, blocked on humans over here. Here's where we're blocked. You know, dev, we need dev for this. Here's where we need you. Dev, Dev. And then, so then, then it just asks me, where would you like me to crank through? And so I set artificial limiters on this thing right now because I don't want it to be fully autonomous because I want to trust, but verify. Now, as I trust the system more and more, I might be able to give it more and more. And because keep in mind, if we're doing this for clients on the signal side, it has to be secure. Right? And so we need a human in the loop here to take responsibility and ultimately do the work. Because we're getting entrusted with doing this work. We're entrusted with strong strategic ideas and strong execution. That's ultimately what we're supposed to do at my company, single grade. But that being said, check out this video over here. If you want to learn the sales agent I'm using right now to book amazing meetings with companies like Google, that's over here.