OMG! Did You See This New OpenClaw Strategy?

I run 6 AI agents. There are 64 cron jobs. They break constantly. They hallucinate all the time. And they cost real money. But they also resurrected $6.6 million in pipeline. They helped me hire superstar engineers. And they also got me invited to speak at a major conference for my multi-trillion-dollar company. Most people build AI agents to do demos to sell you a course. I'm not doing that. We build AI agents to drive revenue. And I'm going to show you exactly how we did that. Let's get into it. So the first one I'm going to share with you is the Deal Reviver. We also call this the Deal Resurrector. I was looking at our HubSpot, and that's our CRM, and I looked at the last, I don't know, 2 years or so. You look at the last 2 years, there are about 229 closed lost deals. These are deals that we lost. And oftentimes when it comes to selling services, people, 99% of people just aren't ready to buy. They're not in market. They might go with a competitor or timing might have changed or something might have happened in company. What we did was we took a look at all closed lost deals that are maybe beyond the last 60 days, the last 90 days or so, and then but less than 2 years or so. What we found was that there were a lot of great companies that fit our ICP that we pitched that we should sometimes just reach back out to. And it's interesting because I was just talking to my podcast co-host about this, and literally they got Heineken just by doing that, just by sending a follow-up, sending follow-ups for the last 4 years or so. That's how they got them back. Back, right? The aging process that we run with this is we have a process of time decay scoring. We look at their probably beyond the last 60 to 90 days or so. And then we're also looking at the last 6 months, last 12 months or so. And then maybe we might go a little beyond that. Even if it's 3, 4 years or so, if you actually pitch them, they might have new contacts in place. And you might just reach out and say, hey, we talked a while back, we pitched you, what do you think about talking? Like, has anything changed? And we actually have revived deals doing this. The trick behind this is if you want an agent to do it for you, I use OpenClaw for this. These deals were literally sitting in the CRM doing nothing. So the second story that I want to share with you is the Beat AI challenge that we put together. Not only that, it helped us create our agent Cyborg, who helps us focus on recruiting. So here's the story. Anthropic put together something called Beat Claude, and they're used to hiring engineers, and I was looking at it, I was like, oh, this Beat Claude thing is pretty interesting. I really like this GitHub repo. So in this case, I will show you what this looks like. So this GitHub repo, this one that I created called Beat Claude, is a fork of Anthropic's version. And this one is very simple. It's the hiring challenge for AI-native marketers. Beat Claude to get an interview. If you want to work at Single Grain, you have to beat this, okay? You can't tie with AI. You can use AI all you want, but if you only tie AI, we're not going to hire you, okay? So that's the whole gist of it. But then these are the roles that we're looking for. So like, a chief operating officer slash GM, a sales director, a head of talent, YouTube strategist, like all these roles over here. Okay, so by the way, if I click into this one over here, you can see there's a brief for the role itself and it gives me a role-play situation. So it's a Series A company, here's attraction so far, here's the market context, and here's your task over here: create a 90-day go-to-market strategy to get them from $0 to $50K in MRR. And then you have no constraints, um, here's scoring criteria. It's all mapped out for you. Here's the expectation for this thing too. And if they're not willing to do this, great. We're going to eliminate them from the hiring process. And I encourage everyone to do this because if they aren't willing to do this, they just weren't willing to— they're just not that into your company. And that's okay because it filters them out and you end up saving a lot of time. Now with Cyborg, it actually— I started to go a step further. I'm like, okay, for this GM sales talent, like we built a dashboard for this. Okay. So, you know, this dashboard over here, we're still working on this right now, but you can see that, you know, we have the email sequences going, you know, what's active, what finished, what's the open rate, what's the reply rate, the click rate on these things, how much are we saving versus the tools that we're paying for right now, you know, what are the numbers of approvals, number of reviews, and you can see the number of meetings that we're having, the number of outreach. And so this is still waiting to get hooked into our warmed email inboxes, and those are still warming up right now. But I just want to show you that this is what our dashboard looks like. And once it gets going, man, sky's the limit, right? Like, There's going to be a lot that we can do with this. This is better than anything, any other recruiting tool that I'm using right now. And so when you start to play around with this stuff more and more, whether you're using OpenClaw, whether you're using Claude Code, I'm using both. I'm combining both and I'm using Claude Code for these deeper builds. I'm using OpenClaw to have the agents talk to me within Slack, to have my agents talk to my team within Slack, to have it talk to me in Telegram as well. I just want to tell you that once someone went through this Beat Claude challenge over here, the very first person that we talked to Ended up being a superstar. And when I got to this person, one of the first questions I like to ask is, how did you prepare for this call? And the very first question he had, he was like, oh, well, I was prepared for this question. I'm like, okay, so what did you do? He's like, one, I read your book. I was like, oh damn, like that's a nice way to inflate my ego. You tell me you read my book, but he did actually read my book. But more importantly, he was able to hang with me when I talked to him about the AI stuff. I really love the fact that, you know, he took a beating as an entrepreneur for the last 2, 3 years or so because you need a lot of resilience to work in this world. We look for people who are unreasonably resilient. And so we ended up, you know, making a job offer. He accepted and he's going to be joining the team. But I'm just telling you, this is the very beginning of this stuff right now. And forget all the demos that you're seeing right now. Forget all the, you know, the fancy, like the newest tools, the workflows, all that kind of stuff. Like you got to think about how can you apply this stuff for your business so you can grow faster. And that's what I'm trying to show you right here. So let's move on to number 3. All right. So number 3 is the story engine. So the way this system works is that there's a cron job that runs at 6:00 AM. So a chronological job. Okay. And it basically scans for content. It ingests the content that I'm putting out. So I have my podcast, obviously I have my YouTube channel, and there might be things that I say on Twitter. It'll ingest my content. Also look at what's happening in the space as well. When I look at my Slack, this is what I'm actively testing right now. We have this channel over here. Okay, so we have this AI content channel over here and you can see how it's calibrated. This is, maybe I'll make a future video on calibration loops. The whole idea here is that, you know, you can see today it's like It basically took a piece of content that we had and it came up with a bunch of different angles for it. Okay, so X article 91, and then it's like it writes it over here. And then here's what you can write for LinkedIn as well. And here's some riff bullets if I want to record a long form from it. So this is a way to repurpose my content into other forms. It takes an atom, which is a content that has like an insight, and then it tries to explode it. Okay, now this isn't necessarily— I would say this is a part of the system, where it gives me atoms, it finds the right moments and things like that. And then we want to repurpose it because I want an end-to-end social media engine right now. What I would say is where it's doing a really good standup job right now is on my, my X articles. Okay. So when we look at my X articles over here, you can see that this article over here, how my OpenClaw creates X posts that average 85,000 views. So this one, 128,000 views. This one, 72,000 over here, 105,000 over here, 89,000 over here. Right. So a lot of this was about OpenClaw, especially when it was hot. And then when I test out these other topics over here, like 1,700 or so, but like this other one, AI infrastructure, 47,000 or so. And this one, I'm talking about something else. So my key point of showing you this is I'm showing you the receipts. This is actually driving more, not just views, but leads ultimately. When I look at the— we have a post over here that I think is this one over here. Okay. This one has 1,163 likes on it. And this one has, I want to say, 44,000 views. Yep, exactly. 44,000 views on Instagram. And this is just a a data, a dot chart. Okay. The cool thing is when I work with my OpenClaw, it knows a lot of the stuff that I'm working on. It knows the types of charts that I like. And so it'll pull, by the way, it'll look at X to see what charts are performing well right now. And it'll give me, it'll try to basically repurpose that chart based on what I'm talking about, or I can dictate what I like. And this morning, actually, it gave me 11 charts. I think we scheduled 11 charts and they were actually really good. I'll give you an example of one that was published this morning. So This one I actually enjoyed because I'm like, okay, I do say this. This is one of the quotes that I said. Money amplifies your nature. AI amplifies your intelligence. Where the world stands on AI adoption. So there's a big dot chart for this that did really well, but I wanted to have a version for me, right? And so, and you can see there's this @EricSiu at the bottom. There's this watermark right here. Money amplifies your nature. AI amplifies your intelligence. Think about that. If you're generous, money makes you more generous. If you're greedy, money makes you greedier. Money doesn't change who you are. It just turns up the volume. Yeah, it works the same for you, but you're free, right? I said this, but I didn't write this caption over here. So I wrote the caption for it here and I put the call to action at the end. This is really cool because it's, it's already at the point where it has access to the Buffer API where it can publish for me. So to social media directly. So whenever I approve it, also as I approve, it learns over time too. So this system compounds. And so you might as well just start building this stuff right now because you're not going to get it on the first day. You have to keep working with it. It's almost like when you're onboarding an employee, it takes a lot of time to get it right. And usually, just by the way, like let's say you're onboarding like a salesperson, it might take them 6 months. Sometimes it even takes them 12 months to get going, right? So the more you spend time with this thing, the more it's going to do better. And right now we're actually piloting these enterprise revenue agents with our clients, but the only reason we're able to move them forward is because I'm testing them on our own stuff right now. And so that's, that's what I would encourage you to do. So, you know, one, we're doing it for clients, like they don't want to spend time doing this. And so I'm showing you what's possible with this so you can, you can get a hang of it, right? And then you have to experiment based on whatever tool stack that, that you're most comfortable with. Real quick, if acquiring customers has been a struggle for you, and you are trying to figure out how AEO works, answer engine optimization, or the new version of SEO, how paid works, how all this AI stuff is going to play into your customer acquisition strategy, then check out my ad agency, Single Grain. It's singlegrain.com. And if it looks like it's a good fit, we'll help you with a free marketing plan. So again, go to singlegrain.com, and we'll see you on the other side. Story number 4 is SEO that self-repairs. And what I mean by that is my Oracle agent basically will scan my Google Analytics, it'll scan my Google Search Console, it'll look at AI search results, and it'll basically detect content decay. It'll look at traffic drops, it'll look at conversion drops as well. And then it'll come up with drafts that we might think about auto-refreshing inside of WordPress and using our SEO product ClickFlow for that. And so we don't need to wait, you know, for quarterly SEO reviews anymore. We can constantly ask questions on what we should be doing. What are the best areas of attack? We have these agents, my OpenCLAW agents that were working with me previously in my Telegram I've had some of them invade my Slack channel, and my team's actually collaborating with these bots. And it's not only helping them get data pulls very quickly in minutes, it's also coming up with insights for them. It's also going back and forth with them, and it's also giving— like, once insights pop up, they tag other people in. So it's like it's spawning a bunch of idea babies that are spreading throughout the organization, and we're able to take action faster because you can actually have these agents even take action for you. And that's still very early days right now, I would say, but it is in fact able to use all the tools that we have, and it Like the very first test, and by the way, it also deleted a bunch of pages that were getting no traffic. And so that actually helps from a crawl budget standpoint for your website, just to make it easier for search engines to figure out what the heck is going on with your website to begin with. And sometimes, sometimes that might lead to more traffic for you and hopefully more conversions. Story number 5 is the way I'm using it as my chief of staff, my CEO operating system. So, you know, when I have my morning briefing, it's going to show me CRM nudges, it's going to show me commitments that I have, my calendar, It's also going to show me the core stats around my business and what's happening. By the way, my agent, my chief of staff agent inserts itself into my leadership chat group and it points out all the stats and it calls out the tough questions that people aren't addressing. And then we should just hit it on right there. I can ask questions around Gong. I can ask questions around, you know, how we're doing with like certain client sheets that we have as well. And also in the evening, it gives me evening brief for the next day. And then again, the goal accountability, I think that comes on Sundays. That's where it asks the pointed questions. And this thing continues to compound as you teach it more. And the cool thing is, as we work with it in Telegram, within Slack, there's a shared brain, there's a shared memory here that helps everything get better in general. It's important to remember the hard truth here. Building agents is very messy. They're going to screw up. The memory can get wonky. You're going to get frustrated sometimes with them. Things are going to break over time as well. And so there's— the reality is really there's a lot of constant debugging. There are hallucinations. There's embarrassing messages that shouldn't have been sent. There's tasks that really never ran, and, and sometimes they'll tell you that at random. Keep that in mind, but the juice is worth the squeeze, right? So I want you to understand that with a human team, maybe what I'm doing right now would take 15 to 20 people or so, you know, limited hours. There's slow cycles. Humans have emotions as well. And I'm not saying replace humans. I'm saying you want your agents to do repetitive tasks, the boring tasks, and you want humans to work on human tasks. Ultimately, you want to hire humans to do human tasks. Robots should do robot tasks. Agents, robots, they will run constantly. They'll never complain. They'll scan thousands of records in seconds and they'll never stop learning. And I would say, even though their memory is a little wonky right now, it's still better than human beings. So right now I see on X, there's reputable people saying that the AI stuff is a waste of time. OpenCLaw is a waste of time. They say this is theater. They say AI agents are useless because at the end of the day, they built like a custom GPT and they called it an agent. But when you really dig into it, the people that are building revenue systems, Those are the people that you want to look at, right? So I would say that the people that are saying AI is useless, there's no ROI right now, I would say it's more of a skill issue. When you say that, you are proving that you haven't dug into it and you are falling behind right now. So that being said, you can check out this video over here on how I replaced a lot of the marketing stuff that we do using Claude Code.