The Startup Ideas Podcast - Grok Bot: make a 1 person company with agents

Episode Date: August 21, 2026

I sit down with Billy Howell for an inside look at a real business that runs on Grok Bot agent teams. Billy tests hundreds of tools for The Rundown, and he uses Grok Bot to run his local newsletter, The Arlington Bagel, which goes to 6,000 readers every Thursday. He walks me through his chief of staff bot, his research and sales agents, and the routines that move work forward while he sleeps. We cover the token math, the four-week plan he follows, and the business models that fit this tool best: newsletters, directories, and Shopify stores. You leave with a setup you can copy for one project this month. Timestamps 00:00 – Intro 02:22 – Grok Bot Overview 03:12 – What makes GrokBot special 06:09 – Pick one project 08:09 – Creating your initial agent team 11:19 – First week sprint with Grok Bot 13:06 – Automated Routines within Grok Bot 15:51 – Agent Org Structure 17:35 – Best Practices for building Agents 19:45 – Newsletter playbook 21:12 – Plugins 22:24 – Newsletter playbook pt 2 25:43 – Sales agent 29:06 – Iterating and Improving Agents 33:34 – Grok Bot business ideas 36:45 – Site stack builder 38:36 – Directories and landing pages 41:42 – One project for a month 42:50 – Closing Thoughts Key Points Keep one project per Grok Bot account, so context stays clean and tokens stay available. Start with a chief of staff bot, and let it audit your business and name the first three agents. Follow the four weeks: build the team, execute, hire and fire, then automate. Ask each agent for a five-line brief: what shipped, what is stuck, what needs you. Move repeatable steps into scripts or make com, and keep agent tokens for high value work. Newsletters and directories give the easiest entry, and the two models feed each other. The #1 tool to find startup ideas/trends - https://www.ideabrowser.com LCA helps Fortune 500s and fast-growing startups build their future - from Warner Music to Fortnite to Dropbox. We turn 'what if' into reality with AI, apps, and next-gen products https://latecheckout.agency/ The Vibe Marketer - Resources for people into vibe marketing/marketing with AI: https://www.thevibemarketer.com/ FIND ME ON SOCIAL X/Twitter: https://twitter.com/gregisenberg Instagram: https://instagram.com/gregisenberg/ LinkedIn: https://www.linkedin.com/in/gisenberg/ FIND BILLY ON SOCIAL: X/Twitter: https://x.com/billyjhowell Instagram: https://www.instagram.com/billyjhowell Tiktok: https://www.tiktok.com/@saasfatigue Youtube: https://www.youtube.com/@billyjhowell

Grockbot is here, and there's a reason why it's going viral. I mean, anyone who's non-technical could spin up these agent teams and actually run a business end-to-end. Today's episode is for anyone who's interested in starting a business, running a business that is completely run by AI agents. And we show you not just big ideas that you can run, small ideas that you can run using Grockbot agent teams that do $2,000, $25,000 a month. I had my friend on Billy Howell come on the podcast and show you exactly how he runs his local newsletter business using Grok Bot agents. So this episode is for anyone who's interested in starting a business that actually has agents running the entire thing. We're going to show you how to use Grokbot. We're going to show you how to spin up these agents.

We're going to show you the best practices. We're going to show you the business ideas that work here. Enjoy the episode. I'll see you at the end. And I can't wait to see what you built. Thank God, Billy Howell is here. And Billy, by the end of this episode, what are people going to learn? You're going to get an inside look at a business that I built and run with Grockbot,

and you're going to come away with how you can set up your own business using Grockbot. Okay. And a little birdie told me that you think Grockbot is like the best place that anyone could run agent teams, right? So you're going to go through the best practices of Grockbot, how you're using it. You're a guy who literally test every single tool. You test hundreds of tools to millions of people with the rundown. You're going to show us how Grockbot works. You can give us the best practices.

You're going to tell us how we can use Grockbot to build businesses and run Asian teams. And by the end of the episode, can you commit that the person that is going to give their time to this episode will have a better understanding than anyone on the internet to run a business with Grockbot? Yeah. 100% because most of the content I see around Grockbot is wrong. They totally don't understand why this tool is powerful. And I'm going to just transmute all that information to people's brains because I'm so excited about this and it's all I think about. All right. Let's do it. Okay. So Grockbot is this new agent team tool. And if you haven't seen it, it looks like this, Greg. Now what we're looking at here is the team that runs my newsletter. My newsletter is called the Arlington Bagel. You can't. can see it here on the hat. Pretty sweet. And it goes out to 6,000 Arlingtonians every Thursday. And I used to have an intern, my cousin Marshall, run this newsletter. He would research it. He would put it together. He would edit it. He would man the email desk. But cousin Marshall,

I had to let him go because Grockbot is just so much better. Sorry, cousin Marshall. What actually happened is he went to work at a summer camp where they don't have internet. So Grockbot kind of saved me. And Greg, have you used GrockBah yet? I have. Okay, cool. How many agents do you have over here?

I have one, two, three, four, five, six. Six. Okay. So the best thing I think about Grockbot is that you have a limited amount of agents. I did the math on this. I think it's about is 76 or 88 possible unique agents? Because when you set them up, you've only got this many shape and color combinations. And you can upload a custom avatar.

But the point is it's all built around constraints. And that's why it's so useful for me. Because if you're like me, you probably use lots of agent tools. You probably have lots of business ideas. and what that turns into is an agent for every single task you think of a chat GPT thread, you'll never come back to a clog co-work project folder that sits there until you remember about it two weeks later. But what I like about Grockbot is you are forced to be mission oriented because it's one thread,

one bot. It's DMs. It's like talking to a teammate. And I actually think that design thing is one of its best selling points. Color and shape is all the human brain needs to differentiate something. So you can easily recognize all your different little teammates here versus like a grayed, just like a gray thread in chat GPT or clod. It's just, I find this tool saves you so much on switching costs and organization.

So right up the bat, we're saving time using this. Am I sounding crazy here? Talking about the design of the agents in Grokbot? Well, you're not. I think a few weeks ago, the buzz was all around Jack Dorsey's Buzz product, which I think was such an interesting product. But I think what, you know, I posted a video about it. And a lot of people were just like, I don't know how to get this started.

You know, it was a lot of the technical people liked it. But what it feels like is, Grockbot is almost like a non-technical version of some of the features of buzz. So if you're non-technical and you want an agent team and you want this almost like Slack-like experience, Grokbot is a good place to start. Now I think the big question that a lot of people have is, and this is what I want to get to by the end of this episode, is it's not cheap. It's, you know, $2, $300 a month.

You know, if, yeah, how can we, yeah, how can we, where you have it right there? How can we make Grockbot pay for itself, 5x, 10x, 20x, 50x, where do we go? Okay, Greg, you're not going to like this, but you're going to have to pick one project to work on with Grockbot. Okay. So the problem that I've found as someone who uses lots of AIT, tools and builds lots of business ideas is you can't you can't you don't have enough tokens to build four businesses as one at once uh you're you're going to go to debt paying for tokens on even worse than that you're going to have context bleed that's going to make all of your agents

less effective so how grok bot works is these guys actually run on the cloud which is different than a lot of agent kind of team products that you use um i don't know well they spend up a computer. That's the difference here, right? Is they spin up virtual machines for your agents? Virtual machine. They all live on one computer in the cloud. I, like, I'm pretty sure on that. I actually haven't thought. But why that's important is if I do my newsletter in here and then I also do, like, managing my Twitter, and then also I'm sorting receipts in my Gmail. That's just going to add context bloat.

that's going to add files to trip them up. It's going to slow you down and it's going to ultimately burn tokens that go towards your limit. I've noticed that the Grockbot usage plan is pretty generous. I think in the first week I used it pretty heavily, but was able to get through week one with like 10% of my usage left. And that's me running one newsletter. So, Greg, like, imagine someone's starting two new businesses and managing their email. You're going to hit that limit. And you can be like, well, I just set $200 on fire.

And you're going to be frustrated. So I think you need to pick. So can you walk us through like, okay, your cousin quits. You need to create an agent team. Like walk us through some of your agents, how people could think about best practices, setting them up. and just give us a sauce. Yeah, ma'am. So when you start, they give you a bunch of preloaded agents that I can't recreate here,

but I picked Chief of Staff from the list. And then because I already had an existing business, so if you have an existing business, give it access to your existing business docs and have it do day one, just like an audit research kind of plan. So I had it look at my notion, my Slack, my Gmail and I said, hey, take stock of this business. And then I need you to tell me the top three agents that you suggest we create first for this team to drive revenue. Right. Our mission is

to drive revenue. So what do we need? And so my chief of staff went, if we go back far enough, I can probably find it. But that's probably not worth doing. It kind of took it. So it took a deep look at everything and it said, okay, we're going to need a research agent. that researches the newsletter items. So it's a local newsletter. And we give you three things to do every weekend and restaurant, new restaurant news, right? So it was like, we need a research agent.

And then chief of staff was like, I'm going to edit it the first issue. We're not going to do an editorial bot yet. So we just started with research, sales, and then the newsletter goes through B-Hive. So it suggested we have a beehive expert agent that just anytime we have a question about can we do this in beehive. Like how does this look in beehive? Stuff like that.

It goes to that. So if your business is really dependent on a platform, say you want to set up a Shopify store, which is something we'll talk about. Definitely having a Shopify dedicated agent is a good idea just so your chief of staff doesn't have to be an expert in everything. So is what I'm hearing the first step is to set up a chief of staff? Set up a chief of staff. You need to figure out how to get your team mission aligned. And so the mission of your business should be to make money. And that's where you need to start.

Have it set up three teammates to start. One of the pitfalls I see people make is having their agents make too many. decisions for them. So for example, I've been struggling with where does our content live when we're writing the newsletter? Does it live in a notion database? Does it live in local files? Do we have it in Google Sheets?

Agent hasn't been able to make a choice for three weeks. So finally, I was just like, we're doing Notion, no more tinkering. Okay. So for some reason, I think that's just a rabbit hole. People get stuck in with AI is shopping around for the perfect solution. At some point, you're the person that has to make the decision, not the agent. And so then I suggest once you build the team, just run the damn ball. Run with that team for a week and see what you can get done.

So if you're starting fresh, use that team to buy a domain, set up your website, help you set up a new Gmail, help you make your first sales list. So think about your chief of staff as the CEO of the business. CEO is not going to be able to run the business if it doesn't know how it works. So I'll give you an example. Say you want to set up an outbound sales agent for your newsletter and sell ad space, sell ad space. Have your chief of staff do that exact task once. And then once it does it and you've reviewed it and it worked, be like, okay, now create an outbound bot that does that thing that you just did. Yeah, it makes sense.

Basically, you have someone, you know, you call it a chief of, staff, I think of it more as like a COO type person, like someone running operations. And you're basically saying like before you go ahead and go crazy, creating a bunch of new agents for every single job to be done that you can think of, the move is basically to perfect it with the chief of staff or the COO, get it to a point where it works. You're happy with the output. Once the output is pretty good, then you're like, okay, I've earned the right to create a another agent that just focuses on that. And then you give, you know, I would imagine it's important to give that particular agent a goal as well.

So with the outbound agent, maybe it's like you need to send X amount of emails or, you know, we want Y amount of visits and stuff like that. And then, you know, we haven't talked about it. But, you know, within Grockbot, there is the concept of essentially a cron job or a weekly you know, something that you can have it do on a daily, a routine, I think it's called a routine on Grockbot where you can have it automated, right? Yeah, that's exactly. That's where I was going to next. So you don't have to do it exactly like this, but this is how I did it.

I built the team. We kind of learned how to fly the plane. No tinkering was the week two of running it with Grock, with Grockbot. I didn't allow myself to create any new agents, not mess around with skills. I mean, maybe add some connections, but it was really about execution. And then week three is when we expanded the team. So we found those gaps. So, okay, we need someone manning the Gmail inbox because chief of staff can't be doing that every day.

We need someone manning the Shopify merch shop. And then comes automation. I mean, you can do automation at the top two after you build the original team. if you have a good idea of how your business should work. But basically what I did was I said to my chief of staff bought in here, I was like, hey, based on how we ran this this week and our agent team, what are some routines we could set up to make progress while I am not working on this? So I want you guys to move the ball forward while I'm sleeping automatically.

And you need routines for that. So it put together a plan. most of these agents have, depending on their role, a daily or weekly or biweekly, end of day brief. So that's a really powerful tool with AI agents. Let me see if I can pull one up. So it's got this instruction here and it's routine. This is the social one. He runs my, he creates my graphics for Instagram and runs the content calendar.

And so the instruction is to look at your current to-dos, what's new, like basically what did you create today? What blockers do you have? And then write a quick five-line report and send it to the chief of staff. I think that you just, five lines is all you need. More than that is you're just going to be burning tokens and context. And then my chief of staff, Bob, will, it'll pull up what's shipped, what's stuck and what needs you.

And so I told it this too. So what's been done, what's blocked, and what do I need to do to unblock? That's awesome. And the reason why it makes sense to have multiple agents and not just giving the chief of staff to do some of these tasks is because you don't want to burn tokens and you don't want to cloud up the context. Is that right? Yeah.

I mean, you need to think about this. like it's a real business and like it's a real team. You wouldn't have the CEO or C-O, let's say, Greg, of a company going in and making Instagram social posts every day. And like a real person, it takes up mental bandwidth for that agent. And that's the idea. And you can click in here and see it messaging other bots and like asking it for updates, which is pretty cool. That is cool. I mean, that's like a, that's what's so cool about all this. That's like a glimpse into like the future of work. where you have all these agents kind of like messaging each other, getting work done, coordinating work together.

And then you're just sitting there, like, kind of updating the North Star and the metrics and the goals and the context. Yeah, yeah, exactly. And so mostly I interface through this chief of staff bot, but if I need to go deeper on something, I cut out the middleman like you would at a real company. I would go right to the social person and be like, hey, we need these graphics. I would give it feedback and be like, can you change the subtitles?

Can you change the background? Just little edits like that. And then you can tell it like, hey, when you're done, update Chief of Staff when we're done here. And it'll go report back. So the Chief of Staff kind of routes tasks around. And theoretically, you know, you could in the future, create a

so, you know, bagel social strategist that's trained on how, like, Akrag Eisenberg does his Instagram posts and says that, like, pretend you're, you know,

at Greg Eisenberg, here's some context. Imagine he was reviewing these posts, what would he say? And then feed that context back to social, bagel social. Is that how you do it?

Yeah. I've been lately having, having it do QA loops. Not so, I haven't done it in Grockbot yet, but in like codex. This is a general agent pattern. So what Greg's talking about is how do I automate feedback for work product that agents create?

And there's this hack basically where you say, okay, once you've done this once, create a panel of subagents for Grockbot. It could be, you know, the chief of staff or just other agents on the team and have them do an adversarial review of your work. do that in three rounds. And that has made me so much progress on stuff that normally I would have to come back and painstakingly review and critique. I get a much more policy.

It goes from like a 50% done output to like 90% done. Yeah. I mean, it also allows the human to be less involved, which is helpful. Yeah. So to be less involved, but you really have to run. You have to be in the weeds with the team for a couple days, a couple weeks before you can run it autonomously. And you do that with the routines, like we said, and you do it with having agents review work that you teach them how to review like you. So a useful thing could be, so say, Greg, like, I gave all this feedback on these posts that it did.

And I could say, okay, take what we did in this thread and turn it into a skill that another agent can use when it's reviewing. So it reviews exactly like Billy, right? That's exactly what you would do. And so if you want to see, do you want to see how the newsletter actually works? Do you want to talk about like the business case? Yeah, I think, yeah, absolutely. Okay. So I do, I do because I think that there's thousands of potential newsletter businesses that could be built and run with Grockbot.

Yeah, thousands for sure. So this is a local newsletter. But I've run newsletters in tech and sports for different clients. And basically all you need to get started is a niche and where to get the best information to put in the newsletter. Right. So if you're really into, I don't know, board games and you know all the best board games subredits. and the best board game review sites and message boards, then you can make a newsletter where you aggregate and editorialize

about new board games every week, right? And so what you would do is you would get a Beehive account. Mine runs all through Beehive. I like to use Notion for laying out my newsletter. And you can connect Grockbot to it. it. It's really easy. By the way, Grockbot does connections way, way better than any other agents. I think it's called plugins, right? On Grockbot? It is called plugins. That's true. Okay, so this is a look at my newsletter in Motion. And so in Grockbot. So sometimes this is a

pro tip. If you don't see a plugin that you want to use, or if, For example, you want to do Shopify. Shopify can, it can like the built-in Grockbot integration or plug-in just is skills for using Shopify. You can't actually do anything in your store. So you'd have to install the Shopify CLI. And so to do that, I would just say, hey, set up this command line tool. And it can use that on its own computer here that you can view it anytime. And that's because Shopify has a CLI, right?

Yes, yeah. But it's pretty intelligent. If you want to connect something, it'll just create like a card in the chat and be like, okay, here's where you put your API key or click this to go sign in with OAuth. It's like really easy to use, very low friction. Okay, so basically what you would do for creating a newsletter with Grockbot

is you would make a list of your sources. It can be a Google sheet, it can be a Notion, and then you have your research bot, go out and look. So I do mine twice a week. We send once a week on Thursdays, and I tell it to go look. And then it fills in these cards in Notion with, okay, this is a new restaurant story. Here's when we should publish it. Here's a short summary. And then if it's an event, here's one they're showing Shrek.

It'll look like this. And I have it set up. So then when I move it to process, there's a make.com automation that writes like the actual blurb for my newsletter, how I want it, which is a little advanced. But you can have Grockbot write the actual like news item for you like in the page and notion. Yeah. Why even have a make.com automation?

So this is like a big agent thing in general is once you've got a workflow that's working for you, you want to get rid of all the ambiguity. So I could have Grockbot write the blurbs every week, and it does a really good job. But if the model changes or, I don't know, just depending on how the bot's feeling, it might write them differently because that's just how it is. They're a black box. And so if you're running something business critical, you want to automate it as much as you can. And that also frees up tokens for you to do more building with Grockbot, right? So you want to build, you want to execute, and then automate.

So when, but you can have Grockbot automate stuff. So I can, and Greg, you're right, I can do a good job having it make an SOP. And I can say, okay, every time you write an article, a news story, run this script that formats it exactly how I want. And Grockbot can run a script on its computer every time. So you could do it that way. And it's just, I don't know, you can do more volume on make.com with like an open AI key for little widget tasks like that. So it's a cost thing.

It is. Yeah. So like with the scale of a newsletter, I could I could waste Grockbot tokens, uh, writing news stories on say like 200 potential news stories. Uh, or I could have Grockbot filter out and just keep the best ones and then have my make.com chat tpt automation, write the little two. sentence blurb. I don't want to waste, it's like a real employee. If you have a high performing employee, I don't want to waste it writing a little two sentence summaries over like every week for a hundred items that I may or may not use. But again, I got here because I stood up

the process and did it manually with Crockbot. First, I had it writing those blurbs every week. And I can give you an example. I keep saying blurbs. But this is what these events that it finds basically turn into. And they're formatted the same every week. There's like emoji, event title, the time, and then a summary. That's the part that I use Open AI for. But Grokbot can do it. Okay.

And then sales. So Grokbot actually found me a sponsor this week. And it's because I connected it to my Gmail and had it just monitor for inbound. And there's something I missed like last week with a local actually. bagel shop owner reached out. And he actually invited me and my wife and we went over there this morning and he gave us free bagels. And we're talking about him sponsoring. And that's because Grockbot found it. And it was like, hey, you should follow up and you should pitch him on this exact ad slot. And that's because I set up this outbound or this inbox monitor and this sales agent.

And the sales agent, what you have to do when you're setting up the business, is, okay, I need you to set up and price everything that we sell. And so in this case, it looked at what ad slots we had and how they were priced. And then it updated them based on the market and the growth of the newsletter. And it made a sales sheet that we could show this local business. That's awesome. That seems like a no-brainer. So I'm just thinking about myself.

It's like I should have like, let's say, I have any media, you know, any media property, you should have like an email, I guess, right? And we should have, for sponsorships, opportunities, reach out to sponsorships at whatever. And then basically what you're saying is you have Grockbot monitoring that inbox, pricing things out, sending messages. And then does that go to your chief of staff eventually that reviews it or it just goes, like, how does the flow work? Yeah. it goes through the chief of staff. The sales bot will say, hey, chief of staff,

we're looking at this potential sponsor. Does this email outbound message make sense? And right now, the chief of staff runs all of those by me. And so I'm kind of being like, okay, put this in Gmail and I'll edit it and send it. And then the idea is it'll just improve over time as I give feedback. I, for something like this, Greg, I believe with cold email, it's quality over quantity. for especially a local newsletter. So that's the thing that's worth your time

to make little tweaks to. And the beauty of Grockbot is it kills that writer's block for starting that outbound message. And so here's the trick. If you're new to doing this kind of outbound outreach for your business, even with AI, you don't want to boil the ocean. So you don't want to make a sales bot and say,

you know, get me every, car dealership in the state of New York and write a message to every sales agent there and set up three subdomains so that my email stays not in spam, right? What you should do is set up an automation that goes on Monday that goes, okay, your job is to add five prospects to our list. And then on Monday, okay, pick the top three prospects and create an ad package custom to them and draft an outbound email and then send it to chief of staff. And you'll make more progress doing that. I promise you than doing 200 outbound emails at a time. What about ideas for the agents?

So one of my one of my Grockbot agents is just an idea agent and his role is to come up with ideas for my other agents, like creative ideas. Because I don't want to be the person always prompting my agents to come up with ideas. Like, how do you think about that? For business ideas or? Like, specific ideas for agents. So, for example, like, yeah, like for the social, for example, bagel social. It's like, you know, I think what makes humans amazing is like, we're,

at least creative humans are going to be like coming up with new ideas, always trying to push it forward. Like, you know, culture changes. Therefore, like, you know, you move social content based on where culture is going. And I find that agents, you know, one of the reasons why AI slop is the thing is if it's under the same loop, it just kind of follows the same loop. And you end up getting that creative and you get product that kind of looks the same. So my question for you is, do you have any ideas around using Grockbot to, come up with creative ideas. My example is I actually went and asked it to create an idea

generator and it works with my other agents. Yeah. So how do we keep it, how do we keep the Grockbot agents out of the loop of not improving? Right? How do we introduce new ideas? I don't. Greg, really? Like, I

why I like Rockbot and how I use it is I only add an agent if we really like needed. I just, I'm so anti- Idea creep and agent creep because that's how you burn tokens. That's how you spin your wheels. That's how you don't make progress. You can do what I did at the beginning. That chief of staff audit, you can run it regularly and say, hey, look at how we worked this week.

What agents can we add? What can we remove? What should we turn into a routine or a skill? You can do that kind of automated improvement. But no, I don't have a good solution for that. Well, you're basically saying like, let the agents cook. Let the, and basically don't. Like, it sounds like what you're saying is, and you've been consistent,

and it sounds like, don't create a lot of agents. Create a small amount of agents that focus on high quality, high value work. Make sure that it meets your bar. Have a chief of staff manage these agents. agents, but only create an agent if it is mission critical. Correct. Yeah. And that's exactly how you should talk to your chief of staff who manages your team.

What else haven't we covered around Grokbot that people need to know about? Well, we talked about the computer a little bit, but it can access files on your computer. So I'm not sure how you, I think it asked permission like one time. and then it's allowed to. You can get it on your phone and then it can run either on this computer or if the app is open and your laptops on,

it can do stuff on your computer. I would say my main takeaway from that is like don't go against the grain. So Grockbot works well on its computer and it can read stuff on yours, but I find myself falling into this trap of, okay, you know, generate a report. And then put it on my desktop

And then put it in Notion. And I don't know if other people are like this, but you end up just working on the system more than the actual output. So I'm kind of like, I tell my agents like, okay, work, create stuff on your computer. And then when it's done, send it to me in chat and put it in Notion. Like when it's done, like after I reviewed it. That way you don't have to be like, hey, that thing we were working on yesterday. day like where is it is it is it on your computer is it on my computer is it in the cloud right you're just creating problems for yourself so i think a lot of people are going to watch listen

to this and really like the whole agent team idea is going to click in their heads yes um i think that there's tons of different business ideas that you can build on top of this you know one of them is local newsletters what else Could people be building? Any type of newsletter doesn't need to be local. Probably would be easier to get the ball rolling on one that's not local. That's a whole different can of worms. Shopify, I've got a separate Grockbot account that I pay the $200 for because I'm so crazy about keeping the contact separate.

And it's just Shopify. And so I'm training it to be to use the Shopify developer tools and to source products. I have a sourcing agent, a developer agent, and like an email agent right now. And I'm still setting those up, but Grockbot does an outstanding job of researching products, sourcing products. And so I'll just, what I would do is say, I have an idea for this product. I have, I really like this company and its product. I want to do something that's in the same category, a similar price point. And it knew to go look at import Yeti,

which can just look at people, Shopify's, and tell you exactly what they're sourcing and what their margins are. It did that on its own. I was kind of blown away, and it made these reports on each product idea and was like, here's my recommendation. Always ask for recommendations, by the way.

Here, like my bulletproof system for testing business ideas with Crockbot is, I want to do this, this, or this. So I've got three ideas. Maybe they're similar. Maybe they're different. Go research each of them. Figure out how I can price it, what the cost is going to be,

and then give me your top three recommendations on how to do the damn thing. And it just works. Your task should not be open-ended when you're doing research, or you're just going to end up doing research for two weeks. That's fair. Okay. So you think, I mean, you think Shopify that there's still opportunity and finding products, finding audiences for those products, and then having agent teams manage that whole process. That's, you know, that's interesting because like, you know, on this podcast, we talk a lot about like digital products, SaaS, vertical AI. Not so much on the Shopify side. So it is, it's just interesting. Yeah, I'm not like an e-com expert.

This is an experiment I'm working on, but I think it's nice to work on something physical because the cost of creating digital things is going to zero. There's no friction there. Actually making something physical, a brand that you can just build over time. I think that's a smart thing to do. And over here, I see you have site builder stack. What do you mean by that? Yeah, so I just have an agent on my other account that has this simple stack of Cloudflare, the CLI Tools.

PorkBun is where I buy domains, so it can log into that on its computer and buy a domain. And then I use this thing called Astro for, it's a framework for creating static sites. And it's connected to my GitHub. And so basically, like, I needed a link for, people to book meetings with me. And I didn't want to pay for Calendly. So I said, I got it set up on this stack. And I said, go embed my three different Google calendars on this. And I was literally sitting on the phone on the beach. And it would send me, okay, here's a mockup of this. And I would tell it, okay, make another branch and do a mock up in this style. And it will come back.

And through Cloudflare, it would have separate developer links that I could review the different styles in. And then when I said, okay, I like that one, it connected it to the domain that I owned. And it went live all from the beach. And I set that up in two or three hours, that stack. And I'll probably use it for a long time, just for simple. I need something to exist on the internet. It doesn't need to be perfect or pretty. So is, you know, is that just like internal tools building internal tools or is there an opportunity to build some of these tools and sell it to other businesses? Yeah. So what I would use this for is landing pages. So yeah, it's kind of an internal tool. If I want to sell, say, a local business on being a sponsor, I could just say,

hey, you know, make a landing page that speaks directly to this customer or directories. So Astro specifically is really good at directories. So if you're interested in the SEO game, just drew through that. This is the stack for testing and iterating. And I would just have a research agent that adds to the directory. So if you want to make a directory of pickleball equipment, I would set up a coding agent that uses this stack and a research agent.

And every day the research agent investigates a new brand and writes a great page on that brand as marked down, and then it hands it to the engineer who posts it. Then you would set up an agent that collects referral links and does outbound to get deals once you have the traffic. And you could rents and repeat that if you're interested in doing kind of the SEO directory game. Yeah. So I've done a few episodes on directories in the past. I'll include the latest one with my friend.

Frey, who does a good job, is kind of explaining why directories matter, how it's a good, like, 10K, 25K, 100K a month business. You know, what is the directory? It's basically just curated data on a website. So you're picking a topic and you're like, here's the best XYZ in this topic. And in an AI world, we actually need directories more than ever because there's just too much information. So when it's curated, it's helpful. Also, the whole like SEO game is changing right now. Like what's being, you know, like people are going to chat GPT and they're going to Google

AI overviews to actually get, you know, to consume content. And they need platforms like that need directories. So I think using Grockbot as the basis for your directory is absolutely brilliant. And the beautiful thing about a directory is it's a once you build a strong directory, it's a foundation and you're getting thousands of visits a month, let's say. And then you can say, you know what, I want to build a Shopify store. Actually, you know what? I want to build a newsletter. Oh, you know what? I actually want to build a SaaS. Or you know what? I want to build a mobile app. So it's actually a really good place for people to start. Yeah. And the cool thing about a directory is it ties in with this, not a local newsletter, but niche newsletter.

So say we're doing that board game idea, you could have it build out like board games that are good for parties could be your niche directory. And you could have a newsletter that talks about that too. You could have this on a subdomain or a different domain than your newsletter. And they tie into each other. So you can monetize by linking people to the directory and have a referral link there. You could do those warm leads from those referral links could become sponsors in your actual newsletter. I think the lowest barrier to entry with Grockbot to start a business is probably newsletter and then directory. I think that stack goes together really nice.

Beautiful. All right. I think we covered what we wanted to cover. Is there anything else you want to leave people with? Yeah. I would do this. Try this for a month. Set up Grockbot for one project and run it like a real team like we went over. And let us, like, let me know, like, tweet at me.

Like, did you make progress doing this? I'd like, this is your sign to pick something that you've been thinking about and pursue it aggressively versus trying a new tool and researching ideas for four weeks. So do this. Build a Spotify or newsletter. or a directory with Grockbot. And I think you'll see what I mean. I haven't been excited about a tool like this. Maybe like all year.

Maybe this is my favorite this year. Yeah. So tweet at me. What's so interesting about this tool is we've seen glimpse of this and other tools, but nothing has been as simple and has just like worked as well as Grockbot. So it's really fun. Also, like, the value that you can actually create is massive. And I'm glad that during this episode we covered it wasn't just about the big ideas, right?

These are, like, little ideas that you can start today that anyone can do. And who knows where that goes? Yeah, maybe that turns into a big idea. I'm sure when Rowan was starting the rundown, he didn't, you know, think it was going to get to millions of subscribers on his newsletter worth millions of dollars. So, you know, you start small, you automate what you can, you keep it, you get a few agents. I love your, I love your framework, build team, week one, execute, no tinkering, week two, week three, hire fire, week four, automate. And do it with one business and just keep iterating and creating value for customers.

I love it. Billy, thank you for coming on. I'll include links for where to follow Billy in the show notes in the description. and I want to hear, and Billy wants to hear, what you know, what you think of Grockbot. So in the comment section, please let us know. I read every single comment and I reply to Mo. So we'd love to see you there. And thanks again, Billy, you're a legend.

Thanks, Greg. It's been a blast.