How I Use Grok Bot to Run Marketing Every Day

Eric Siu built a library of specialized AI bots that handle repetitive marketing work like SEO drafting, recruiting, and deal sourcing, freeing his team for high-value tasks. He connects them to existing tools via a single API gateway and emphasizes human oversight to prevent mistakes. The bots follow the QDSSA framework: Question, Delete, Simplify, Speed, Automate.
Chapters
What percent of your work is done by AI today? If it's not above 95% as an engineer, you're out. If it's not above 70% as a non-engineer, you're out. Everyone's going to become an agent manager. It actually means you're going to have to hire more people because anybody on the team that knows how to use this stuff, they're constantly becoming more valuable. They're letting the robots do robot work and then the humans are doing human work. We're going to be covering how we're using Grokbot from a marketing standpoint, some of the learnings that we've had already, how we're using it from a business standpoint. There's a Grokbot library here that we'll go into a little more detail. Look at this. We got Emotion God here, Trial Reels, AEO, SEO Bot. calling it Autonomous Engine Optimization, Mixie. No clue what that is, right? And this is basically full kimono. I'm going to show you guys what's happening over here. I'm going to share some learnings from the GrokBot team as well in terms of what's coming down the pipeline too, because this is hot right now. So everyone using GrokBot right now, you know, everyone likes to talk about the Chief of Staff and it's helpful. The Chief of Staff bot here is kind of the overarching bot that can see what's happening with the other ones and it will kind of route messages to you and tell you, hey, this needs some attention right here. And one of the biggest challenges I have when it comes to AI right now, it is actually progressing work. It's advancing work, right? And So the good thing with GrokBot is that you can continue to have it advance work for you. You can set the routines, and I find it to be the most reliable of the different harnesses that I use right now. So I do use Hermes. I still love Hermes, right? I use Codex quite a bit as well, but both Hermes and Codex fail quite a bit when it comes to setting up routines. Like we have a Loops Creator skill that allows you to make these loops where people talk about these self-improving loops. But when I use it, it's a little finicky. I would just say that. GrokBot has been more reliable when it comes to having routines run. It kind of just works, right? It's easy to set up out of the box. It's easy to kind of share your bots. It's easy to convert skills into bots. I think there's like an Australian ad that says, you know, it does what it says on the tin, right? And that's what GrokBot is. I'm not saying, do you go out there and replace your Hermes? No. Am I saying replace your Codex or your Claude Code? No. Am I saying replace your OpenCLAW? Well, yes, because OpenCLAW sucks right now. So yes, replace OpenCLAW. Guys, I will share these with you afterwards because you guys can just fork these. We have the first mate here. Okay. I love the first mate. It tells me what's wrong with my GitHub in terms of what needs to be cleaned up. It also makes it competitive in our group channels inside of Slack now, because what it does is it sends a message out to the entire team and it says, hey, here's who's actually contributing to repos right now. Here's, here's who's shipping. Here's who's reviewing right now. And here's who isn't, right? And that makes it public accountability. And that's how our organization works now. And it works well. So we have a talent bot over here. What the talent bot does is it will look at inbound talent that's coming in through our applicant tracking system. We have this autonomous engine optimization bot, AEO SEO bot here, and it leverages a handful of connectors that will help you grow your traffic. There's a short form scaler as well, because we want to, we want to scale our short form content more. You can see that GrokBot actually does a fairly good job of editing, but you're going to have to continue to tweak this over and over. We have the X long form writer. That's a skill that we have that's been helping us write X long form articles that have led to us getting a couple million views a month on X. Sponsorship deal desk, this handles sponsorships automatically. I had a lot of these living inside of Codex or Hermes. And I just said, hey, Hermes Codex, I want you to just create handoff MD files. And so I can make these into a Grok bot. And that has worked. If you're watching this right now, you're more than likely paying for a couple of different tools, right? So for example, we pay for LinkedIn Recruiter, we pay for JuiceBox, which is, these are actually agents that handle outreach and they do a pretty good job. And who knows if we're going to continue to pay for them, but as of right now, since I pay for it, I might as well activate it. So it's within your best interest to look at all the tools that you're paying for right now. You have Gong, you have Granola, you have whatever recruiting tools that you have, and then figure out how GrokBot can just run them for you because GrokBot actually has computer use. It can operate your web app, whatever web app that you're paying for through a browser, right? And the cool thing is I just hand this off to my team afterwards. I just give them the templates and then they're rolling with it, right? Because ultimately what you should be trying to do as a, as a leader, whether you're a marketing leader or a business leader, You should be thinking about how do I create this leverage, not just for myself, but for other people. Creating the leverage just for yourself is a very selfish thing to do. If you have a Cursor Teams account, you can actually have a shared account with all of your skills in there. You can share bots in there, and then you can have different, you know, permissioning roles set up in there too. So let's go to the AEO bot. This bot mines 40 real SEO AEO candidates every weekday, not every week. Because we don't have time for every week. We need every weekday now, right? Actually, why is it every weekday? It should be every day, right? That's my bad. We have 7 days out of the week. Why would you not have your agents working 7 days a week? Do they need a weekend? No. As long as we have compute and electricity, they don't need a weekend, right? Every weekday from your search, your tape, you pick 20, ClickFlow drafts them unpublished, and then we measure the clicks, the free consults, and AI share of voice. Nothing goes live without an exact yes. That part is very important. What you should be doing, and actually we had someone on our team face this issue today. So their Grok bot just started to YOLO, right? They connected it to all these things and it started sending emails and like they tried to send it to Sean, but it sent it to the wrong Sean and it just started sending whatever emails, which means it can damage your brand at the end of the day. So I made this little human gate skill over here. Use before any external send, post, publish, delete, spending, or CRM calendar, right? Enforces draft only until an exact current approval. You're trying to get more search visibility, right? You want more share of voice. You want to drive more, like your SEO traffic's going down, but you want to rank better on the answer engines. Okay. We get it. Well, why do you want to rank higher? Do you want to get more traffic? You want to get more leads? If you think about all the content sources, the pillar content sources that you can pull from, or the content spikes, think about all the customer conversations that you have. Think about all the sales conversations that you have. Think about all the internal conversations you have. All the time you spend, spend in meetings every single week. How many content spikes are you cranking out? You yourself, right? How many smart things are you saying? You're probably saying at least 7 smart things a week. If you say 1 smart thing a day, maybe it's 7. Maybe you're saying 5 smart things a day, right? But Those things should all be converted into content. Now, all of you might be saying, well, well, I don't have the time for that. Well, great. That's why you have something like this, this AEO SEO Bot. In the old days of doing SEO, I would have had to go to my content team and say, hey, I think these ideas are good. We'd have to debate these ideas. We'd have to, you know, take a look at Ahrefs, take a look at the volume, and then someone's going to have to do the research here. Uh, you know, is it going to be top of the funnel, middle of the funnel, bottom of the funnel? Is it net new content? Is it an update? Is it a delete? Is it consolidate? You don't have to worry about that anymore. Well, even though there's not keyword volume for this right now, I know that it's a trending topic, right? How do I know it's a trending topic? Well, because I just hang out on X all day and I just, I can tell that GrokBot's a trending topic. So that means that we should attack this from multiple different angles. GrokBot for marketing, GrokBot for SEO, GrokBot for short-form content, GrokBot for business, right? GrokBot for LinkedIn, GrokBot, like GrokBot for anything. I've been talking about agent-led growth on my YouTube, my podcast, and that is just, to me, it's a new version of product-led growth, right? Because you're trying to optimize for agents. How did it know that these are both updates? Well, because it's connected to my Google Search Console, it's connected to my Google Analytics. Wait a minute, but Grokbot doesn't allow you to connect to those right now, right? Or does it? You can use ClickFlow, which is our AEO SEO tool, which has the connectors in there, GSC, Google Analytics. And then ClickFlow can also, I think it has data for SEO in there. So it can spot search trends and AI visibility as well. Anything that has an API or MCP, you should, you should be able to plug into it. So you should be fine with Google Search Console and Google Analytics. So the point is, You want to collect your, connect your different AEO SEO tools. Okay. And then you want to connect some, some form of a content generator. All right. And then afterwards you want to make sure you still, that you have a human in a loop generate the content. So we do have evals tied into this in terms of what we expect out of our content in terms of the templates that we have for different sections on our website. I'm also constantly adding to this bot too. Okay. So yesterday I saw that, uh, you know, someone wrote a post on Amplitude and Amplitude has a lot of problem pages that they create. Ahead of a search term getting more traffic, right? So they've done a really good job with that. I'm like, okay, well, hey, how can we use this? Are we going to make a net new bot? Like, what should we do exactly? This is all I do, right? If I see something nice on X, I see someone share a cool post, or I should see them share a cool bot. I just say, hey, is this an update for us? Or is this a net new bot? And I just shared that post. Okay, pulling that post now. Okay, this is upgrade, not a new bot. Okay. See, you see how short I am here? I'm just like, hey, just, just do this, right? Just, just do this and tell me what you're doing next. Right. Okay. Let's test the upgrade. And then it's like, okay, to convert, is this going to, to create more problem pages for us or give me a list of 10? And it's like, okay, here's, here's the top 10 over here. And then I'm like, hey, how can you verify that these are trending upwards? So this is how I go about building the bot. And this is how I pull from different resources. We still use WordPress as a CMS. I know, haha, it's a little old now, but we use WP Vibe, which is a great MCP. And then we can put that, set that to draft. Notice I'm not saying, hey, set it to just create content on demand because I don't want to just— what if all of a sudden it creates 5,000 pieces of content that are total crap? I don't want to have to clean that all up. We use a tool called Overlap, and Overlap is really great for short-form content. What I'm testing over here is I'm just constantly, I'm having it cut, but I'm also giving it feedback consistently, right? And then honestly, this should just become an eval. Again, the eval is your definition of done. Because when you think about Instagram Reels right now, Instagram Reels has the trial Reels feature, and And then now there's an editing app where it allows you to make 5 different versions of the reel that you make. When you use the trial reels feature, you can actually, if you edit the overlay or you edit the text, or you maybe you cut it a little tighter, that's going to increase retention. And sometimes when you make some of these small changes, it's like A/B testing for your reels. And by the way, when you make your reels, that to me is you're, you're kind of testing your creatives, right? From a, from a paid ad standpoint. By the way, if you're struggling with your AEO and your Reddit right now, that is the number one thing that we've been helping our clients with. So when it comes to Reddit, How do we find the right threads that are ranking well for the keywords that you want to be ranking for? How do we insert ourselves into the conversation but insert ourselves tastefully while not just asking people to buy our stuff? How do we reach out to the right people as well to get included from a PR standpoint, from an AEO standpoint? Also, how do we scale quality content from an AEO standpoint? And that's something we use when we use our software ClickFlow and we manage it for our clients, right? So on a single grain side, we have Reddit, which ties in with our AEO. And we're getting a lot of requests. We're helping a lot of people with AEO right now. So if that's something that you need help with, just go to singlegrain.com. That's with the G, singlegrain.com. And we'll see you on the other side. When you needed to hire someone for YouTube packaging, MrBeast, for example, pays $10,000 for a thumbnail, right? Because each of his videos gets, you know, tens of millions, hundreds of millions of views. Let's say you don't have that budget. And before, we would hire thumbnails people, and each thumbnail would be fairly expensive. So what we did was we actually ended up making a thumbnail skill. Okay, so I haven't converted this into a GrokBot skill just because I think Codex is a good harness for this. This is the video that we threw up on my YouTube channel yesterday. It's already at 10,000 views or so, which is good for my channel, right? Like if any video gets 20,000 views or so, I'm happy with it because I know it's the relevant views ideally that will drive leads. This is what the thumbnails look like. Okay. I believe it's this one that went live and it has 7.2% for leveling up. It's like a black background and there's like purple outlines. Okay. And more or less it's me. Okay. Mostly me. Right. And I wanted the GrokBot logo and I'm like this, I'm like, This, this is pretty good. Now, is it world-class? No, but it's pretty good. 10 bots, 1 boss. It might be. I think this one's actually better, right? In fact, I feel compelled to use this one now because it looks better. But my point in saying this is that I continue to refine this YouTube packaging skill, which by the way is on the Skills Dojo. I think the Skills Dojo has an MCP too. So we used to have to pay, I don't know, like anywhere from $50 to $100 to maybe $150 for a thumbnail or so. But now That cost is gone. 2, we don't even need to wait for these things to get up. And so that GrokBot video, I recorded that on Sunday just because I was, I had inspiration hit me. And besides the editing, the packaging was already all done. And so once we get this to a point with the editing bots doing a good job with overlap, if I really want to get a video out today, I can do that, right? Based on our styles, because we know what we like and we can just show like 3 videos to the AI and say, hey, This is what success looks like. And let's make that into an eval. This is our Skills Dojo, literally skillsdojo.com/templates is where you can go. You can get these templates that I'm talking about, right? So this SEO, the AEO SEO bot that I was just talking about, it's right here. Pre-call intelligence, read-only sales prep for founders and revenue teams. This one's helpful. Sponsorship deal desk. Those of you that are getting outreach for sponsorships, we get like 20 or 30 of these per week. This outreach agent is really awesome too. So what this does is it'll look through our CRM, it'll look through closed lost deals, okay, and try to revive them with a new angle. All right. Or it'll look for deals that have stalled out. So maybe deals that were moving along that need, need, need a little nudging. Not only that, these outreach agents will also find lookalikes for people that have reached out to you, let's say within the last 12 months or so, that you have deemed ICP fit and you want to find more people like them and you can reach out with a different offer. Connect it to your LinkedIn Sales Navigator and you don't even need an API key because you can just run it through. Your Sales Navigator and a Mixie mix panel and analysis bot for usage analytics. Damn, that's pretty cool. I didn't even know he had this, right? In fact, this Mixie should be publishing to our Slack channel on the product side every single week so we can see how each product is doing. Hey, look at this. Submit a bot. You guys can submit your bots in here too, and we'll consider adding them, right? So we have the Skills Dojo. Like we've, we've just today, we added this bot section for bot templates. You can go get them. And then the Skills Dojo, ultimately you can add whatever skills you want. You can, you can download them, you can fork them, and you can make them into bots as well. You want your team to feel good about it first. You want your team to actually show usage from this first before you try to put more tentacles into or embed this thing more into your organization. When I first started using OpenClaw and Hermes, I tried to make one mega agent. Inevitably it would break, right? Because you feed it too much context. There's too much. It starts to get drift. And what happens is that you have to reset the bot. And that's happened to me with OpenClaw once or twice already. Don't build one giant agent. Ideally each bot has one clear job. And then from there you can figure out, okay, once you feel good about it, then you can make it into a template and then you can hand it off to other people. You can put it off into the wild or you can hand it to your team. So you don't need to make 30 bots at once. Just try to make, I don't know, 2 or 3. If you're getting the most value from your creatives right now, and a lot of the creatives are coming in from Reels, from UGC, maybe you do that first. If you're getting a lot of out— a lot of success from outreach, great, do that. If you're getting a lot of success from AEO SEO, do that. Do more of what works initially. Once you nail it. Then you scale it and then you can make new bots, right? You don't have to try to do everything under the sun. What is the most annoying thing that you're doing every single day? What do you hate doing? One of the most annoying things for me is if I spend 30 minutes making a YouTube video and our packaging sucks, that just drives me crazy, right? And so I took it upon myself to actually spend the time to make that skill. I had a sponsorship routine running in Codex, but it was break all the time, right? And so now that I have GrokBot running it for me, it's, it's auto responding to people now that anybody that wants to do a sponsorship, it's like, It's going to be $15,000. And they start negotiating, right? And then the bot just keeps responding, it's going to be $15,000. It's going to be $15,000, just over and over and over, right? And so that way you don't have to deal with it. So think about the most annoying thing that you would like to get off your plate today and think about how you can skillify it. And I guarantee you, you're going to be a lot happier and then you're just going to try to do more and more and more. And that's how this goes. A working bot needs an ICP. It needs safety limits, output caps, and evals, right? So Again, like I said earlier, you don't want to just say, go make more content, go make more high quality content. You want to say create 4 candidates for approval based on our writer guidelines, based on content spikes from the content ingester, and then measure them after 7, 14, 30, and 60 days and only look at the last 7 days of content, right? If you like the output, then you can just make that into a routine that repeats every week. I think the problem right now with a lot of people is that, you know, they, they just try to automate out the gate and I would just encourage you whether you like Elon or not. To use the Elon algorithm. Whenever you are building something, okay, you would think that, oh, because of AI, you would want to automate immediately. Nope, you don't want to do that. So there's 5 points here. QDSSA, QDSSA, right? Think of it like that. The first step is you want to question, okay? So whatever the requirements are, you want to question, okay? Then you want to delete. You want delete steps from the process. Then you want to simplify, right? Then you speed up. And only then do you automate. Let's use Tesla. Okay. As an example. So when they were running into production issues, they couldn't produce the cars quickly enough. Well, why was that? They looked at their factory and they're like, how do we just cut down like half of these robots? Like, because there's so many moving parts, literally. And one of the engineers said, hey, let me think about this. And the next day he came into the office and he brought in one of these little matchbox cars, right? And he showed the bottom of the car. He's like, look, there's only one part that connects everything here. And that one change put Tesla ahead by 5 to 7 years when it came to manufacturing cars. And now I think a lot of cars are, uh, car companies are doing it, but that was because they questioned the requirements and then they deleted a lot of these steps and then they simplified, then they sped up, then they automated it. Right. And that's how we should be thinking about these bots in general, right? We don't want to just automate out the gate. Let's nail it as human beings first, and then we can try to scale it. One of the most important things when it comes to AI and growing your revenues right now is having a connector. So we have this API gateway and this allows you to connect to the most important tools out there. It could be ChatGPT, it can be Google Search Console, it can be all these other tools that you might have, but they aren't talking to each other right now. And you want to make sure that your connectors obviously have the right permissioning, there's the right security setup, and all these different things, right? And so we have combined that all into an API gateway, which we're calling the Single Brain Gateway. And if you want to get access to that, you can see all these different tools that we have over here. You can add your HubSpot, your Gong, and we're continuing to add more connectors in here. And that's going to allow you to scale the things that you want to do. You can plug it into different agents. This is what we use with our agents, with our revenue brain, with our company brain that allows us to not only grow ourselves faster, but to grow our clients faster as well. So if you want to check it out, just go to gateway.singlebrain.com to learn more, and we'll see you on the other side. I see us making more bots, continually maintaining them, updating them, and hiring more people that know how to manage agents. That's why when we hire people now with the talent bot that we have, we reach out to them, they say yes on LinkedIn or whatever. We'll send them a list of 6 questions and then you kind of just go down the list there. So we have an eval, it has all the success metrics that I just talked about over here, and then it'll evaluate all of them and then it'll move them to the next step. Don't think that you can train these people. I've tried it a lot in the beginning of this year, and we've just realized that when we just raised the bar higher, it's worked a lot better from a marketing standpoint and from a business standpoint. This is the best AI stack for business today and for marketing today. So you should be using Codex. Codex is the best harness right now, and it's your daily driver for building things, getting things done. And it does a good job with YouTube thumbnails, for example, does a good job with editing. It does a fairly good job with writing. Grokbot should be your daily driver for having reliable routines done. So for example, could be reaching out to people, reviving dead leads, or handling sponsorships and looking through your email inbox, right? I found Grokbot to be the most reliable and easy to set up, and it's cool to share the templates out there. And then number 3 would be Hermes. Hermes is great because I can use it on my phone pretty quickly, on my desktop pretty quickly, um, and it has the most context around my business because I've constantly been building with it over time, and it's open source and I can kind of do whatever I want with it, right? And so when you combine the 3 together, you have these superpowers. I don't think one's— you shouldn't just handicap yourself and say one's better than the other. I think, you know, the answer is multiple ones as long as they're reliable. And then obviously if you're using WhisperFlow, you're talking to these things, you're going to get things done a lot faster. Loom videos too are great. I can send it to Claude or I can send it to Codex and I can say make this into a skill, or I can go Grokbot and make this into a skill. This one's free, right? The Skills Dojo that I was talking about. So skillsdojo.com to upload your and download the best Grokbot templates, and then you fork them and make them into your own. So Hermes is in play every day. We still use Grok every day. I still use Codex. So on the right side of my screen over here, so my other screen, I have Codex wide open. Grokbot is open on my left screen over here, and then Hermes is open with the Hermes desktop app. And I also have the Hermex mobile app, which is kind of like the desktop app, but for mobile, because they don't have a native mobile app yet. Those are my daily drivers. I still open up Claude every now and then. I think Claude Design is really good as well. And even though we pay for Claude Teams for the team, my team, my company, I don't really use Claude. And because I think Codex is the best harness right now. What I mentioned earlier is GrokBot is much easier to set up and it's more reliable when it comes to running routines. When it comes to running routines with like Codex or Hermes, they tend to fall apart and they tend to fail. And then whenever you check back in with them, oh yeah, yeah, sorry, I failed, right? But like, they didn't even tell you that they failed. I'm like, hey, did you just fail on this? Like, oh yeah, yeah, yeah, I failed. And so I get really frustrated at times with Codex and people are like, oh yeah, you know, you can just /goal and, you know, just run loops. No, we're not quite there yet, right? Maybe you can do that for engineering because it works a lot more verifiable. But when it comes to work that we're doing around marketing, don't think that you can just get away and type /goal for everything. You got to make sure, hey, whatever you're trying to do from a marketing standpoint, maybe you're trying to lift your conversion rate from 1% to 1.5% in the next 8 weeks or so. That's verifiable. Or maybe you want 100% of your webpages to average 50 milliseconds. That's verifiable. But when you say, I want you to just grow my product, that's not very verifiable, right? So that, that's, uh, you put in, uh, you know, poor inputs, you get bad outputs at the end of the day. GrokBot as a harness tends to be very short. And I think this is just how GrokBot is used to communicating. It doesn't do the best job of writing long-form content. When we have this as a skill and it runs in Hermes or runs inside of Codex, the output is a lot better. Claude is probably still best for writing, maybe not Opus, but maybe for Fable. But that depends on you if you want to burn the tokens for that. GrokBot is going to become multiplayer. So the agents can talk to each other already, the bots can talk to each other, but they're going to add in the ability for you to add your teammates in there as well. All right. And they're adding— they're going to be adding in Slack too, where the agents can pop into Slack and you can interact with the agents inside of Slack. And so we're already working like that a little bit with Hermes inside of our Slack. And it's been good because people can kind of get the context that they need and just ask questions. The next step is where this is more or less secure, right? You know, you can have different permissioning levels, you know, you can have different security governance, you know, all that set up and you're gonna get more team collaboration. So that is coming. That's coming down the pipeline. So Skills matter, evals matter. Eventually, once loops work really well, loops matter as well. Um, but—










