The Go Local Brief

Go Local Brief Pilot Episode: 0.0
===

[00:00:00] This is the Go Local Brief, a podcast from Go Local Interactive, a digital and technology marketing agency based in Kansas City with clients all over the United States. We operate in the self-storage, multi-family, home services, and senior living industries. On the Go local brief, we'll explore marketing, news, technology, insights, and much, much more.

Let's get to the show.

Shane: Hey, welcome to the Go Local, brief . I'm Shane Adams. I'm the Director of marketing, and with me as always is Jason Barrett, the CEO.

Jason: Hello everyone.

Shane: We're gonna talk about marketing technology.

We're gonna talk about SEO, paid media and all sorts of things. Let's get started. Cool. Let's get started. Hey, Jason, you're really excited about Open Claw, and I am. I think it's a really interesting technology that kind of combines like agentic ai. What's your take on it and where do you see it going?

Jason: Yeah, it's interesting because it's, it's the first time in a long time that I've gotten wildly excited about something. And the [00:01:00] reason is. Um, you know, it, it basically takes what was there with open AI or Claude or something like that. And it wraps, uh, more automation and ag agentic help around it. And it maintains memory, kind of maintains state.

There's something called, compacting that happens in a conversation. So there's only so many tokens, that are allowed inside of a conversation window. And a token would be like, think of it as , some chunk of information that is given to the chat window. Okay? And the bigger the context that can be handled by a model, the better.

But they can only take in so much at one time. Whether,

Shane: wait, does that mean that it's better with bigger prompts or

Jason: it handles more information all at once? So it's like if you said, Hey, here's, here are my handwritten notes, just a couple little bullet points.

That's small amount of tokens. Or here's like the entire Encyclopedia Britannica. Now [00:02:00] let's talk. So the one is a lot more token rich than the other.

Shane: Mm-hmm.

Jason: Or like video versus text. Or audio versus text. Text is very small.

Shane: Right.

Jason: But the more of it, the more tokens that are required.

Shane: Okay.

Jason: Right. Okay.

So it's got at its fingertips, you know, all this information that's built into the model, but then you're also giving it information in the prompt that you want it to do something with in relation to that model.

Shane: Hmm.

Jason: So. Hey, here is our forecast and budget for 2026. Look through it, and then I'm, I'm going to ask some questions about it, or I want a presentation created, or whatever these things might be.

Mm-hmm. Is a lot more, uh, context than if I just said, Hey, can you take a peek at this? Uh, copy and paste an email that I got from Shane a little while ago. How should I respond? Right. That's very simple.

Shane: Right.

Jason: So, um. The other thing is I could, I could have a run on conversation with my open claw [00:03:00] instance all day long, and it's remembering everything we've talked about all day long and until it compacts.

It's all still in there.

Shane: Oh, okay.

Jason: Like I could say, Hey, remember when we were in the, uh, executive meeting this morning and we were making jokes about mm-hmm. Tom's son. Yep. I'd be like, yeah, I remember that. Now you get it too far back past the compacting point.

Shane: Okay.

Jason: And it, when it compacts, it's like, it's a whole new conversation.

Shane: Mm-hmm.

Jason: The thing that Open Claw does, that's different than, um. Just the straight model.

Shane: Mm-hmm.

Jason: Is it will try to write some of that out to files that are called markdown files.

Shane: Oh, okay.

Jason: So it'll put those in like a memory Mark md or it could put some information in it. Sold md It could actually, my setup, I have it doing, um, every single day.

Is another markdown file so that it can go back. I can go, we talked about this last Tuesday.

Shane: Mm.

Jason: And it can reference that markdown file and go, yep, that's right. We did. Okay. I [00:04:00] see the, I see the file that you were talking about. I see the image you were talking about, whatever that might be. So that's what's different than if you, if you, do you use chat GPT.

Oh yeah. Absolutely. So you know how there's conversational threads.

Shane: Yeah. Yeah.

Jason: It's keeping track of all of it at one time. So you don't have to go find the thread where you were talking about, you know, pillars of marketing.

Shane: Okay. Okay.

Jason: And, and restart that conversation. You can just be in the same, the one window and go, Hey, we were talking about pillars of marketing.

Stereo Mix: Oh,

Shane: okay. Okay.

Jason: And it would go, yep. Got it. And then, hey, I need a, um, I need to send Jason an email and it will already know because, oh, that's Jason at Go Local. Yep. He's. In this role, he's gonna want this information, that sort of thing. It's already got all of that.

Shane: Okay. And, and so, the difference between because 'cause chat stores memories in the, like in a similar way or is it different like the way.

I know that like, uh, on the free one, for example mm-hmm. You can run outta memory, right? Like that, just like you're, you're running out of, is that Exactly, you're running outta the markdown tokens?

Jason: Well, yeah. Yeah. You're [00:05:00] running out of the what it can handle.

Shane: Okay.

Jason: And it says you need to start a new thread.

Shane: Right.

Jason: And then if you've ever used like, um, Claude Code or, or maybe you use an IDE development environment like Cursor or something like that. It'll go along so far, and then it'll tell you compacting. It'll say compacting, and then dot, dot do. And you're waiting and you're waiting and then it comes back.

Some of those ides can, then they're, they're inside of the development environment's, inside of a project, so it already has awareness of its environment, which means, Hey, we're working on this website, we're working on this script, we're working, whatever that might be. It wakes back up. In there and goes, oh, I, I I've got this.

Yeah. I see. Um, whereas when you do that in through the free version on the web mm-hmm. It's a whole new ball game. You're, it's like, Hey, how are you? What can I help you with?

Shane: Right.

Jason: First time.

Shane: And it's like, yeah. It's like I don't, I don't know you.

Jason: Yeah.

Shane: Who are you again?

Jason: And it's trying to build, they're, they are doing a good job of building some of that back into those windows.

Um, [00:06:00] and if you remember, we could all, we could create GPTs Uhhuh. You can still do that. Um, and that, that basically gives it preset prompting and pre a little bit of preset context. You could drop files in there to say, you're gonna need these. Right. We're talking about Go local. It's the Go Local handbook.

It's the, you know, our, our format, our template for our, um, presentations or whatever. You're gonna need these and then now we talk.

Shane: Right.

Jason: You know, but with the open claw set up, it's already got that built in every single time you're talking.

Shane: So, uh, what, what what has been interesting is, I think over the last couple of months, you, you've kind of gone from like being super all in on N eight n

Jason: Yeah.

Shane: To yeah. This open claw. Is that because you feel like open claw is like the better of the world, or what would you, like, what, what would you, what would you say to the, the difference between those two?

Jason: It's interesting. N8N is really good at structured workflow. Okay. These steps happen and they happen this order all the time, [00:07:00] or maybe there's some sort of if then in there, but it's not, it's not, it doesn't vary wildly.

Shane: Okay.

Jason: And when it does, you, you could tend to use, um, you know, like a Google sheet or something like that for tracking or for variance or maybe like direction, but it always goes from point A through Z

Shane: mm-hmm.

Jason: In a certain order.

Shane: Right.

Jason: You can, you can route down through there based on. Um, certain,

Shane: yeah.

There's some rules or what? There some rules. Like you can say, okay, if this happens, go here. But, but, but ultimately you're trying to end up at Z

Jason: Right? So with open claw, it's just got the intelligence to go, oh, I, I got this. I, I need to do these things. Mm-hmm. You don't have to like, you don't have to give it the exact recipe from A to Z like all the time.

Mm-hmm. It can, it can come up with that in intelligence on its own. Mm. To go, oh, I got it.

Shane: Yeah.

Jason: Secondarily. Innate end fires off of some sort of a trigger. There's always a triggering, um, [00:08:00]

Shane: oh

Jason: yeah. Instance or situation that happens that that kicks it off,

Shane: right?

Jason: Whether that be Salesforce like a closed one on a, on a prospect.

Or, um, a ticket gets updated or you can send an email. It, it can do any number. You can hit a button, you can fill out a form. Any number of things can trigger it,

Shane: but time, period, whatever.

Jason: Yeah. Yeah. And the same thing can happen, uh, with an open claw set up, but it also has like, um, what would be, it's called a heartbeat.

So it literally kind of wakes itself up. That's the other thing that is, is unique about Open Claw is it basically wakes up that. Uh, prompting window to say, do we have anything we need to do? No. Okay. Back to sleep. Do we have anything we need to do? No. And it's sitting there thinking and working and acting for you or for us.

Shane: Mm-hmm.

Jason: All the time. Um, and so you can also then schedule out what would be we, we've, uh, developers have always called them CR jobs, chronological or chronology or, you know, timed jobs, [00:09:00] which are. You know, basically what pops up on your phone when it's like 10 minutes to the podcast meeting with Shane.

Shane: Mm-hmm.

Jason: That's a cr job running on my phone and pop 10 minutes out, you've got a meeting with Shane? Yep. And I can dismiss it. Remind me in five. Right? So five minutes, pa pop it, you know, same thing can happen with, uh, open claw, where you can say, Hey, re look at my schedule, look at my calendar. Any meeting that I have.

Prepare any kind of, um, information I would need prior to that meeting. Get it ready for me the morning of

Shane: Okay.

Jason: And email it to me.

Shane: Nice.

Jason: So it, it then would schedule out those crons or on the heartbeat would wake up and look, or you could say, Hey, remind me tomorrow night, very much like you do with like Alexa or something.

Like, Hey, remind me tomorrow night that I wanna watch such and such a show, or I need to send my mom a, a happy birthday text. Yeah. Or something like that. Mm-hmm. And so you can literally just say. Schedule this and it will do that.

Shane: Right.

Jason: Change the schedule, you know, whatever you might want to do.

Shane: So, [00:10:00] I mean, you've shown me what open claw can do, and I think it's like, even just like the, the surface stuff that you've shown me is, is incredible.

Jason: Yeah.

Shane: Uh, how do you see it kind of, uh, I I, it feels like we're at a tipping point, right? Like we're kind of like at this like place where. This technology like went, like it, it 10 Xed itself like overnight almost. Yeah. And uh, how do you see that impacting, uh, the way we work?

Jason: I think for us it's going to be like, oh, I should have also said this thing can call tools.

Um, and for us it's going to be like a big tool that we can use to do better work for our clients. You know, our people do so much analysis. They get down into the data. They, they, they, they're trying to do so much for so many clients all at once, which, you know, they do a fantastic job of that. But the system can go into, um, that data and, and, and [00:11:00] surface.

Information back to the team that they can enact on in a much more, oh, I don't know, like automated fashion.

Shane: Mm-hmm.

Jason: You know, so that it can be, I think it takes our people from like 10 to like 15 or 20 in terms of what their capabilities can be like uhhuh. With the same 100 people, you can do two to three times the, the work and the quality increases.

Um, you know, the level of intelligence increases. Uh, it, it just gets, everything gets smarter and more interconnected. So we've, we've tried really hard to make sure that paid search is aware of what's going on in organic. Who's aware of what's going on over with the content, the website, everybody's talking?

Well, this is one of those things where. The, the, the agents might work their way through the website, surface opportunities from an organic perspective that then it can go and pull GA for Google Analytics and our ad spend and go, oh, based on, for example, we're [00:12:00] overspending or we're being really pressed for our cost per clicks over here in this particular area.

It just can spin up additional content or suggest that. Um, and then, you know, hey, we need to rank like this. We need to work on our schema like that. I've gone ahead and done it. It's out there ready for your review. That sort of thing. Like the, it's just taking a, a holistic look at everything and, and, and surfacing it.

Shane: That's, uh, I mean it's, it's pretty incredible what the, what is, what is able to be accomplished with, you know, as long as you know how to manipulate it. Right? Yeah. Like, and I think that's the thing that I like. I think there's a lot of folks who are, who are like, oh, like they're kind of scared of it, right?

Yeah. Like they're, they're like, uh, hey, this is, this is gonna take away my job, or, right. And the thing I, I always kind of come back to is it, it won't take your job if you, if you at least know how to use it.

Jason: Yeah.

Shane: Right. Like if you know how to manipulate it and to control it in a way that gives [00:13:00] you, we're gonna need folks who have the ability to look at a, a tool like this and go.

Okay, I can do this now. Yes. Because I, I now have this extra set of hands, essentially. Yes. Doing things that I wasn't able to do.

Jason: It's like everybody having their very own like, uh, like personal assistant or like an extremely good intern or research analyst or data analyst or, you know, that's right there at their fingertips.

Mm-hmm. And they, you know, it doesn't necessarily know what to go and look at. You can train that into the system. Um, it, it needs guidance. Mm-hmm. And it certainly needs, um, approval and like the human in the loop going, like, you just hallucinated. Like, that's, that's not,

Shane: yeah.

Jason: Uh, you got that wrong. Let's hear here to look.

Try it like this.

Shane: Right.

Jason: Um, you know, you're way high on your math. Like, for example, like I see a lot of, um, Hey, if you did this, you'd make millions, but y [00:14:00] no, you know? No, you would not. Yeah, that'd be great. Yeah. But there's a lot more that goes into it than just that. And so, um, as it gets smarter, some of that stuff's worked out of the system, but it really brings humans wanna work with humans.

Shane: Mm-hmm.

Jason: Like what, what I think is really fair is to go. Some of these tools are fantastic. Right. But I want to talk to you.

Shane: Yeah.

Jason: I don't want a chat bot. I don't wanna interact with that chat bot. Maybe there are some times when that's acceptable.

Shane: Right.

Jason: But I have a real question and I wanna talk to, to some person.

Shane: Mm-hmm.

Jason: You know, and it, it, I keep hearing the term AI slop and I think that's pretty, uh, it's fair. You know?

Shane: Yeah. Well, I, I think, uh, was it, was it Affleck? Uh, I, I think Affleck was on a, he was on a podcast and, and he and, and Matt Damon were Yeah. Were pitching the Yeah.

Netflix, uh, show that they Yeah. That they just did. And he said, he said something like, AI is designed to be average.

Jason: [00:15:00] Yeah.

Shane: And it's true.

Jason: Yes,

Shane: it is true. However, it, I mean, average across. A whole lot of things

Jason: Right.

Shane: Gives you a bigger toolbox. Right. Whereas like yes. Are are you gonna get, you're not going to, Claude isn't able to write the Great Gatsby

Jason: no.

Shane: Right? Like it can consume it and maybe summarize it. Right. But to be able to actually execute the way that Yeah. Escot Fitzgerald did. Yeah. It's not gonna do that.

Jason: Yeah.

Shane: And I think that's, I think that's what you're getting at.

Jason: That's exactly it. I think humans are so. You see it, you know it. Like, you can look at an image and know, oh, that's not real.

You certainly, when you put that image into motion with video and you go, that's definitely not real.

Shane: Yeah.

Jason: And then there's something about text You go, I like, I know a, a person did not write that.

Shane: Yeah.

Jason: You know, and there's something to, um, like the creator of Open Claw said he misses typos. Like, I miss seeing human typos because, [00:16:00] and I thought that was really.

Like, how interesting is that? Because everything's perfect. Right? Too. Too perfect.

Shane: Yes, yes.

Jason: You know, and you're like, this is not human.

Shane: Right?

Jason: And I don't know, there there are, there's a time and a place for that. Like, look at these numbers, get the, get the analysis, right. What do you think we ought to do?

Build a plan, document this project, all of that stuff. But as for how a human, the human would go, ah, you don't really need that, you know, or Right. Yeah. That's, that's, that's, that's extra, or this would be a big deal if you could get this worked into the system, you know, that sort of thing where they, there's just something about people that makes our output, um, you know, it's gonna be a pre, at a premium.

I do think that the AI generated cont there's a, there's a place for it for sure. Mm-hmm. And I, I don't think we can deny that. And I think as it gets better and better and better and better. There will be some aspects to our lives where we would, we would, [00:17:00] I think we'll look at it and go, oh, you let a human do that?

Like, for example, maybe like reading an X-ray and trying to find

Shane: Oh yeah.

Jason: Maybe. Or an MRI, like I know that, that I'm, what I'm saying

Shane: is, yeah. Is that an art form or is it a, or is it science? Right.

Jason: You know, if, if, if the, if the, uh, error rate is extremely low.

Shane: Yeah.

Jason: I think you'd rather have. Yeah. AI take a peek at it, you know, and go, no, that's okay.

Or no, you've got a real issue right there. You might've been, it might've been overlooked or,

Shane: yeah.

Jason: Second opinion.

Shane: I was working on, I was working on something last week where, uh, I was, I was trying to work out a system.

Jason: Yeah.

Shane: Uh, for case studies.

Jason: Yeah.

Shane: And I was like, this is what I'm thinking about case studies right now.

Jason: Mm-hmm.

Shane: And give me feedback on this thing.

Jason: Mm-hmm.

Shane: And, and, and I think, I think I was using chat, I can't remember if it was chat or Claude, but I fed it into chat and chat basically kicked back and it was like, oh, I [00:18:00] get this because I understand systems.

Jason: Yeah.

Shane: And that I think is like, that, that delineation made me think about it very differently.

Yes. Almost. Because you go, oh, okay. Systems make sense. Art rule based

Jason: systems.

Shane: Art is not a system. Art

Jason: breaks the system every time.

Shane: Right? It does. And and, and now,

Jason: not every

Shane: time. Not every time. Because the, the, the thing that I keep, that, that made, uh, something you said, made me think of was that, uh, the, the, the test of.

Video that was like, this is where AI is right now. And it was like, will Smith eating pasta.

Jason: Okay.

Shane: And, the distance it has come in three, like two or three years. Yeah. To where like, the first one looks like AI , really, obviously. Yeah. And the next one, well, our eyes have been trained to see it as ai, but it's way less than we thought it was.

Yeah. Yep. Right. And that I think is where, okay. It, it, it's not art.

Jason: No,

Shane: but I can describe that scene Yep. And go, I, yeah, I could. [00:19:00] That's a system that I can describe.

Jason: Well, and I think that's like the, the, the role of a director in make, in a, in, in a movie.

Shane: Mm-hmm.

Jason: Right. Is what you're going to see with our roles in the business.

Shane: Hmm.

Jason: You're gonna have that vision. I, I want the camera in tight or I want it back. I wanna use the wide angle I want. This to happen from the far right and below and end of the screen. I want somebody to walk the now, you know, like mm-hmm. That sort of human, um, direct directing is what's going to be in there for, Hey, I need a website.

It needs to do this. Don't do that. Like, let's monitor the, the user experience. We, we will make adjustments on the fly, these colors, this banner, um, run paid, but land them on that page. Do we, you know, you're, I need an analysis of content gaps,

Shane: right?

Jason: Based on usage, based on some of the best, um, [00:20:00] sites that we see online, et cetera.

I need to know, is our schema updated? I need, you're gonna just all the time being, you know, taking in and then directing out.

Shane: Yep.

Jason: And I think that's the way that we're gonna see open claw go. Mm-hmm. Or, or, or, or whatever it turns out to be, you know? The, the creator has now been hired at OpenAI.

Shane: Oh, sure.

Jason: Yeah, of

Shane: course.

Jason: And so I, he wants the project to, to, to remain open source out there. But I, you, you, you can anticipate that this age agentic world will be. Available through each of the, the major, you know, uh, Google.

Shane: Yeah.

Jason: With Gemini, you'll see something from Anthropic. You certainly are about to see it from open ai.

Shane: Yeah.

Jason: Mark Zuckerberg tried to get him at Meta.

Shane: Sure.

Jason: They were in the, they were in the hunt, and so everybody's going, ah, that is it. Mm-hmm. You know, we need that. We, we, we want something like that.

Shane: Well, and I, I agentic even this just a term is something I feel like just. [00:21:00] Like magically came on the scene in 2026 and it is like February.

Right? Right. Like, you're like, wait, where did this come from? Yes. It felt, it felt very quick. Yeah. Uh, and, and open claw I think was driving so much of that. Yeah. But I also think we could see some of the, some of the, I felt like Claude, especially with Claude Code, was really kind of pushing that forward.

Uh, but man, it's, it's a new world.

Jason: Well, it's funny because, um. So I'm a huge Bitcoin fan. I certainly don't like the price where it is right now, but um. So Bitcoin was kind of created by, uh, somebody named Satoshi. Right. And he basically took he or she or they took all these pieces and put it together.

They were already there for the most part. It's the same thing here with Open Claw. You had models, you had, um, tool calls from models. You had, um, CR jobs are certainly not new.

Shane: Mm-hmm.

Jason: Uh, markdown files are basically just, um. Ways [00:22:00] to basically take notes in a structured format.

Shane: Mm-hmm.

Jason: And just kind of put all of those things together.

And I'm oversimplifying, there's more to it than that. But that happened in like November or December of 25, released kind of quietly. Right. And then some people started finding out about it over the, uh, Christmas break. That's. Not uncommon with college students being out, uh, professors, people like that, that even us, you know, where we take a little break between the, uh, end of the year and, and the start of the next one.

And, and so that's when all of a sudden you started hearing these whispers about, uh, it was Claude Bot.

Stereo Mix: Mm-hmm.

Jason: You know, and now open Claw and, um, excuse me. When we went from, you know. Are they bots, are they agents? It was vibe coating.

Shane: Yep.

Jason: All through 25 and into the, in part of the year, last part of the year.

And the cool thing you would do with vibe coding before we got to this situation was if you, one, you'd be doing it two, if you were really a power [00:23:00] user, you'd have all kinds of different windows up. So like you, you, you would have cursor up with six different, um, theoretical agents running at any one time.

Shane: Okay.

Jason: Well. Then along comes, you know, what's now open claw? And you can, you, you can spin up as many agents as you want, as many agents as your machine can handle as many agents as you've got tokens. Uh, you've, you've gotta buy these tokens with subscriptions to either open AI or Claude or Gemini or whoever this might be.

And. Like in my instance, I have a personal instance running and I've got I think 48 or 49 agents at my disposal at any one time. They're not all running at once, but if we wanted to have a party, I can get 'em all going. I can get 'em all spun up and go, Hey, we've got a brand new, we, we need to write new software.

I want marketing material prepared for it. Give me a website developed and, and next thing you know, I literally have to tell it. Remember, you have 48 threads. Uh, [00:24:00] in our local processor, save five for our conversation, and you can run to go to town with the other 43 mm and it'll just,

Shane: just go to town.

Jason: Go to town.

And my, my primary contact, my primary, um, agent. We will spin up the rest of them. Oh, wow. So I don't have to talk

Shane: to You're not, you're not prompting the rest of 'em. You're just telling the one

Jason: and it's,

Shane: and it goes, oh, I know that I'm gonna need this thing. And it goes and it gets it.

Jason: Think very much like here.

If I, if I were to come into the executive meeting and say, uh, we're, we're, we're, we're moving into a new space. We're gonna go into, um, swimming, pool marketing or something like that. Hmm. That would mean for you, I need, what are we going to do from the marketing perspective? I need from, um, the finance department.

What, what does this look like in terms of projections from, uh, net revenue side and, and a cost side? You're gonna need to get those numbers, uh, finance department from the sales team mm-hmm. In terms of what they think we might be able to accomplish.

Shane: Right.

Jason: And so I say [00:25:00] it, that goes out to the. Primary team who then pushes it out to individual contributors that, that they might not need information from, then it surfaces back.

Uh, to me it's the same scenario that that is working inside of that machine.

Shane: That's crazy. Yeah. Yeah. That's really cool. So

Jason: yeah.

Ryan: Thanks for joining us for this pilot episode of the Go Local Brief. Stay tuned for more marketing and technology insights from Go Local Interactive.