![04 | Jake George Talks AI Workflows and Solutions for SMBs at Synthorial Labs Artwork](https://www.buzzsprout.com/rails/active_storage/representations/redirect/eyJfcmFpbHMiOnsibWVzc2FnZSI6IkJBaHBCR1Y1clFjPSIsImV4cCI6bnVsbCwicHVyIjoiYmxvYl9pZCJ9fQ==--ee1e991aa908687767fd12344f531f1e2bb9c79e/eyJfcmFpbHMiOnsibWVzc2FnZSI6IkJBaDdDVG9MWm05eWJXRjBPZ2hxY0djNkUzSmxjMmw2WlY5MGIxOW1hV3hzV3docEFsZ0NhUUpZQW5zR09nbGpjbTl3T2d0alpXNTBjbVU2Q25OaGRtVnlld1k2REhGMVlXeHBkSGxwUVRvUVkyOXNiM1Z5YzNCaFkyVkpJZ2x6Y21kaUJqb0dSVlE9IiwiZXhwIjpudWxsLCJwdXIiOiJ2YXJpYXRpb24ifX0=--1924d851274c06c8fa0acdfeffb43489fc4a7fcc/Jake%20George%20(1).png)
The AI Marketer's Playbook
The AI Marketer's Playbook is an actionable podcast focusing on AI and marketing. Each episode covers AI strategies, tools, and trends that are changing marketing. Listen to interviews with industry experts, analyze case studies, and get practical tips. This podcast is for anyone looking to leverage AI in marketing to improve results.
Tune in to stay updated on AI marketing.
The AI Marketer's Playbook
04 | Jake George Talks AI Workflows and Solutions for SMBs at Synthorial Labs
In this episode of the AI Marketers Playbook, host Audrey Chia sits down with Jake George, founder of Synthorial Labs, to discuss how bespoke AI tools and intelligent agents can transform small to medium-sized businesses. Jake shares insights on automating workflows to scale operations without increasing overhead costs. They explore the benefits of using AI for internal and external tasks, from managing inventories to automating customer communications.
Tune in to learn practical steps for leveraging AI and automation in your business, and discover Jake's journey to AI innovation.
Join my weekly Newsletter: https://lp.closewithcopy.co/welcome
Welcome back to the AI Marketers Playbook, where we cover actionable frameworks to help you leverage AI in business. I'm Audrey Chia, your host for today. And today I have with me the AI genie and my good friend, Jake George. Now, Jake is the founder of Synthorial Labs, where they create bespoke AI tools. And intelligent agents for small to medium sized businesses to help them automate their workflows and scale without adding extra overhead costs. Welcome to the show, Jake.
Jake George:Hey, thanks for having me on Audrey. I'm glad that we could arrange this.
Audrey:Super cool. So for those of you who don't know Jake and I are partners, we've been working together for quite a number of things and he is a brilliant mind. And for those of you who don't know him, Jake, tell us a bit more about yourself.
Jake George:Yeah, so what I do at my agency, Synthoria Labs, we help like, like you said, small to medium sized businesses implement intelligent AI agents and automated workflows, to kind of like help them scale without just hiring on more employees to do all those manual tasks. And as we've done, together, sometimes we'll even develop like white labeled platforms for other companies to help even their clients, outside of themselves. So a lot of times it's like, we might develop a solution for them. Some, similar to the, LinkedIn launchpad and then they're like, Oh, wow, we get great use out of this. Can we also offer this, to our clients as well? So we do, mainly focus on like. Internal and external AI agents. So things, you know, for internal, that would look like it could be like inventory management, it might manage their inbox, help them schedule meetings, create content. you know, there's a million things that you can do, especially with custom development and our team of developers. you know, there's no job too big for us. And then externally. we help them like automate their communication lines. So if you're running, say a business where you use, you get, let's say, tons of incoming calls and most of your sales are done over Instagram and Facebook. So then you could have a external agent that helps you automate the messaging, close your leads, schedule appointments, send more information, qualify leads. you know, really depends on like the type of business there, but it's like there's two ways to implement different types of agents, you know, the internal or more external facing.
Audrey:Yeah, so let's talk about that, right? I think a lot of people, you know, think AI is this like huge amazing thing, right? But they don't really know how to apply it in their business. So if I am a small business owner and I'm just starting out, right? How do I apply, for example, like what you said, the internal agents and how do I apply an external agent?
Jake George:Yeah. So if you're just starting off, I mean, to be honest, they're not all like very useful agents are kind of difficult to build. Like it's not the easiest thing. so if you're using stuff internally, and you want to learn that, and it's like, you want to really get into automations and use it for yourself. I would recommend learning a few platforms like make. com. Zapier, relevance. and then just know how to use the open AI API, which I know for a lot of people's like, you know, if it's, if you're not so tech savvy, it's like, you'll probably have a better time just using like, and no code platforms and then build automations based off of that. So that could be as something as simple as something like, Oh, maybe, you have a ton of leads and you're, you're, you're a solopreneur or something like that. And you have to manage all of those in your CRM, update the pipeline, using low code and no code tools. You can integrate with, a lot of popular CRMs and then you can just connect like an open AI step to that kind of the way that you do that with no code tools is that they have like these visual modules where essentially they just give you text inputs and they handle everything on the back end. So what you could do is go into something like make and then you could have it like, okay, when I, you know, send records. It splits them up this way. It splits up like they're in first name, last name, their email, whatever other information you want to put in. And then it sends that off to your CRM. So you can make something like as simple as that, using AI. I mean, everyone's pretty familiar with GPTs that I personally think suck, but a lot of people like them. I just think they're not, that's kind of on the end of like, there's like different parts of AI. You know, there's people that are like. It's just chat GPT. They think they, they use it like Google. They're just like, give me some cool marketing ideas. And then they're like, that sucks. I could have thought of that on my own. And it's like, yeah, probably you could have. So, you know, then you can take it a step further and do like advanced prompting, which you'll get better answers, but then you can even take it a step further than that and have it start like automating actions. So that's where you, that's an automation. That's the difference between just using AI. And automating with AI is connecting it to the apps and tools that you use, having it make decisions and then take actions. So, you know, like I said with the CRM, it's like you can make it just do basic actions based on a text input. So rather than just giving you an answer, you can say like, do this thing and it does that thing. And then you take it even a step further, which is, you know, very complex for most people to understand. And then it's like, You know, you're developing like little, like micro SAS platforms for these companies or like their own, you know, web apps and stuff like that, where essentially it's like, you don't even, we'll take it to the point where it's like, they don't even really need to talk to it necessarily. It's just taking action. So for example, it's, we work with, have a project, it's, like a commodity trader. And so when emails are incoming, it's able to like scan the PDF documents and like read out different parts of data and then log that into a, essentially it's like an Excel sheet, log it into a database and track all of that stuff. So you don't sit there and tell it every time, like, Hey, do this. Hey, do that. Cause then you might as well do it yourself. So really like the end goal of automation is something that you don't even really see running. Like I can't, you know, I can't even really show you it. Because it's like, it's not like you go and you just like activate it. And then a magic thing happens. It's like doing these little background tasks that would normally take, you know, teams or yourself, you know, like 10, 20 hours a week of just repetitive, monotonous tasks that you can train an AI to do.
Audrey:Definitely, I love what you just said. So Jake shared so many golden nuggets, right? I think this is also where you as a user, start learning how to use AI. So for me, I came from a non technical background. And when I first started, it was just chat GPT prompting. And like what Jake said, I created GPTs, but the results were not consistent. and it was frustrating to some sense, which is also why I reached out, and then we found a better solution for that. And I think the second step to consider is also what Jake mentioned about using AI and automation together. This is something that a lot of businesses are still not fully leveraging right now. a lot of people are still dabbling in one or the other, or just even not trying either of it, which is a huge waste of opportunity if you ask me. But yeah, I think that's super interesting. And Jake, I would love to know, like, what was your background before AI? What were you doing before this?
Jake George:So before this, I, I had another agency and we did like crypto and NFT marketing. so it wasn't necessarily heavy into development, but we did have a, developer team, because if you are familiar with like blockchain and how it works, I mean, it's all like, It's built on like smart contracts and you know, it's code written by developers and stuff. So it's like I had had experience, like I wasn't the one developing them, but building a development team. And so from there, again, if you follow, you know, blockchain and Bitcoin a lot, you know, it went like chew in 2020 and then. So it's like, you know, when you're up here, people are willing to invest a lot more in development and marketing when you're down here, people not really want to invest. So we kind of just, it was sort of like a natural pivot as like, you know, GPT 3. 5 had just come out. so then it's like, because before that, it's like you could access like LLMs, you know, there were still like, there was an older version of chat GVT, but it's like, they just didn't really get that. So they weren't really smart enough to the point that it was going to be super useful. For the types of solutions that we develop. I mean, I, I know it's, you know, it's been used in like a lot of stuff that you don't even think about, even like on your phone and like, you know, predictive texts and things like that. It's like, there's different, you know, little parts of where you're kind of always been using AI, but for the things that we developed, you know, once a GPT 3. 5 came out, I just kind of, you know, became obsessed with it, like I was still doing the. blockchain stuff back then, but I just spent hours and hours a day just like researching, like using it, trying out things and just kind of realized that I was like, you know, I was using it just as like most people start off where it's just like, you just have it, you know, like say create content for you or give you ideas or you ask it questions and it just kind of was like a light bulb in my head and I was like, Wow, this is so useful. And then just seeing how that would integrate into, because another thing, like, obviously software and automation has been around for years and years and years and years. It's another thing that it's like, you know, you don't even notice you're using a lot of the time. Like there's automations and every single tool that you use, but then realizing how it's like, you can connect all of them and put these AI steps in there, which can essentially just replace the manual work of, you know, Oh, I have to go here, read this part from this email, go put it into the CRM and then write another email to reply. And then, you know, log that data somewhere else, export as PDF, like whatever you got to do with it. And it's just all those like annoying little steps that, you know, will typically take up hours of your day without you even really noticing. Like you might not think that you're, you know. If you sit back and think about like, Oh, what could I automate? You know, there might be like one, two things that come to mind, but it's like, there's all these little like mini micro tasks that you do throughout the day. So I kind of started using it for myself. And so I would just like automate those for fun. and then it's like, you know, I just had so much fun building those out. I'm like, Oh, we could build like way more in depth stuff. And so, yeah, then we sort of pivoted from there.
Audrey:Yeah, like what Jake said, right? I think one thing that most people don't realize is the time it actually, you have to take in order to master a skillset. So you mentioned how you actually spend time like playing around and getting interested in experimenting, right? I think it was pretty much the same for myself. when I first started. I really had no idea what I was doing. I was starting off with one liner prompts on ChatGPC. I'm like, Oh, that is amazing. And then I started evolving and learning this new skillset. But like any other skillset, you need to take time to do it, invest in resources, or reach out to a partner who can support you because that's the way that you supercharge that learning process, and really start seeing the light. The results in terms of productivity gains or efficiencies, but Jake, you also mentioned something about abusing agents, right? Can you tell us more about how you would maybe construct different agents to support a company or business?
Jake George:Yeah, so there's, I mean, there's several ways to do that depending on what the, what the agent actually does, but essentially it's like what the, the main, part of intelligent agents is that it's like you give them tools. So it's like these tools can be coded by developers. You can make them in low code or no code apps, but essentially what it is is like you're building out tools to complete tasks based on decisions and input from an AI and then you give it access to all of those tools, which it's, it's similar to GPTs in a way, but on a more like advanced level, because it's like, When you have, developers, it's like you can make any tool that you want. And then it's just having an agent to run it and know how to put the input in. So really, I mean, the, the possibilities are endless. Like it can integrate with like. Slack, Trello, Gmail, Facebook, geez, what else? Like a notion, you know, it can, you can connect it to like your business phone number and have it do phone calls. You can connect it to your Twilio and have it manage your text messages. And those can be like inputs, like communication channels to have it take other actions as well. So it's like, for example, one of the projects we've done is like an inventory management system for a media company. So they have tons of cameras, tons of batteries, tons of drones, and it kind of gets overwhelming for them to know where all that stuff is at any given point. But where they kind of keep track of like their projects and stuff is on Trello, and where they do their communications is Slack. So giving an agent access to a database that it can control, and it has a full inventory of what they have, and it can tell from their Trello, What projects they're working on, what stage they're at and what gear needs to be checked in and out by who. So it's one of those things that kind of runs in the background and if they want to make any like specific edits to it where they're like, Hey, it's not on the Trello, but I'm just taking a camera for a day to go do this shoot real quick, well then they can just communicate with it. So really, I mean, I could go on and on. Like there's a million things that agents could do. Essentially. It's like, Anything that, say, like a personal assistant would do for you, an AI could do. And another, one of my favorite things, and this was something really mind blowing when I first saw it, was there's this tool called AutoGen. So essentially what that does, it was, it was one of the first ones that I had used, it takes agents. So agents that are connected to tools and it puts them like in a group chat together. So what you can do is you can make these agents and now a lot of other platforms have started integrating this as well. So you could have one agent that's like your project manager. So let's say that's connected to like your Trello and your CRM. So it kind of oversees, it knows what needs to be done. And then you have one that's your content writer and the reason for splitting them up, because you might think like, well, why don't I just make one super smart agent and write all the different prompts of what to do? It's like, Theoretically, sure, it could work, but another thing you have to keep in mind is that even with the, like, token limits and the amount of, like, prompts that you can give to agents, specifically, like, most people are aware, it's like, with ChatGPT, you can only give it so much text and stuff, there is a limit. But even if you completely fill that up, it doesn't mean it's going to be 100 percent accurate with your prompt. And that's an issue that I think a lot of people run into with mega prompts and stuff. They just think longer is better. Like, if I just keep adding information, it's going to be smarter. And it's like, yes, to a point. You want to have, you want to have good context. You want to have good examples. You want it to be a clearly and concisely written prompt. You want to break it up and make sure that you have like, you know, role task format examples, you know, it's like you've seen all of the frameworks, that people do in LinkedIn posts. So they're very good. But if you just overload one agent. And try and make it like a know everything sort of agent. It's just going to mess up on the little tasks. If you want it to do 10 different things, it's just going to mess up. But the cool thing is it's like when you use agents to call other agents, so you have like a central, like main. Project manager agent and that essentially just knows it's like just a project manager. It knows who needs to do what and then each agent that it can talk to has, access to tools that they need to do their job. And also a prompt that is, you know, Shorter and more concise than if you mix them all together to only do a specific job and then they all communicate So that's how it's like say that you have like one inbox manager and then one database manager that also Tracks manages a CRM and then one that creates content, you know So it's like rather than trying to make an all in one sort of thing, you just have the main agent that you communicate with and it delegates the tasks out. They use the tools to do their job and then send the reply back to the project manager, that can determine if it's good or not.
Audrey:That is so smart. It's like having an AI team leader with mini AI subordinates and then you are the boss and just sitting there and be like, hey, get my, my stuff done, man. Wow, that's super cool. I think, I think like what you say also is about like prompting, right? Not overloading one agent with too much information. Even in copywriting, I've also seen the same thing. sometimes I see people creating mega prompts for landing pages. But I've tested those output, right? And I compared it to perhaps breaking it down into different steps. And you would realize that the output is pretty good. very significantly because I think it's also like not overwhelming the AI with too much information and I have a feeling that they may not know really what to index on so when you have like a mega prompt they might miss out on certain things or might you know over and overly emphasize on one part of the prompt and then you would see that the output differs significantly so that's something. but having said that, Jake, is there any kind of demo you can show us with regards to your agents or anything you're working on? I think it would be really cool to take a look at it.
Jake George:Yeah, for sure. I can pull up, like, one of my, multi channel agents here just to kind of give you, like, a basic demo of how that could work across, like, Let's see, like Facebook, Instagram, and my phone. let me just see.
Audrey:Yeah, and for those of you who are listening, for those of you who are listening, you might not be able to see the screen, but I think Jake can voice over and walk us through what we are seeing right now. And of course, don't be afraid to just come on to our YouTube channel to take a look at the full demo. Cause, I think it's going to blow your mind. Okay, over to you, Jake.
Jake George:Okay, cool. I will, share my screen here. let's see. So I'll just try sharing the window. I know it's going to do this infinite mirror for a second, but, okay. Let me if you can see this screen.
Audrey:Yes, we can. Awesome! Cool,
Jake George:so this is, this is where I have my demo. So, what I could do is, I'll just like ask it basic questions about, my company. Just say like, what sorts of workflows does Synthoria Labs automate? And we'll give it a second here because it's going to then go into its memory and then like look up the information as well. And then it will simply message us back right here as well. So just give it a second to process.
Audrey:And, and for those of you who cannot see what's on screen, we are on Jake's company's Facebook page, right? And we're in the DM function. Of the company's Facebook page. And he just popped the message there.
Jake George:Cool. So it says, since Oriel Labs automates a variety of workflows, including customer service tasks, sales funnels to attract ideal customers, client onboarding processes, and document generation. these automated workflows help streamline business processes and improve efficiency. So now what I'm going to do is I'm just going to say. My name is Trevor. Don't forget that. And I'll show you one of the capabilities of being able to, use like a multi channel approach. So, got it. So now it remembers my name. So now I'm going to go over to my Instagram now. Let me know if you can see this.
Audrey:Yes, we are on your Instagram, Peach.
Jake George:And I'll just say, what's my name? So with this, what you can do is on the backend. So there it goes. It says, your name is Trevor. So it's like, it remembers this, like it knows who I am across all these platforms. And so that's one of the benefits of having a connection to either like a database or a CRM. is that it can search that information as long as, like, you have the person's contact already in there. If it's a new person, obviously they have no message history. But it's like you can look them up and then just, you know, keep the communication consistent across all channels. So it's like you're actually talking to, like, a real person that knows who you are, rather than every time it goes, Hello, how can I help you? and restarts here. So now, I'll do one more thing. I'll say, Can you give me a call? So what it's going to do here, it should then ask my number, so that I can input it and then it's going to call me. So I'm just going to give that there. And then this should pop up on my phone in just a second. I will pull it up and hold it to the mic. Cause it will be, might be a little bit hard to see my screen, but here, you're able to hear that.
Audrey:I heard something really, really soft. Oh, okay. Yes. I hear it now. Oh,
Jake George:there. I was just wondering what you guys do.
AI:Software development company. Focused on helping businesses automate their sales funnels. We create custom tools that eliminate the need for multiple tools, and integrate them with the sprint. We also offer consulting services to understand your needs and design solutions around those inefficiencies.
Jake George:So you kind of get the picture there. you know, we can go on and on with that, but essentially it's just like, you know, you can give it all the context of, oop, now it's going to a mirror here, but we can give it all the context of everything that, you know, was said in the conversation. You can request a callback. You could have that. I mean, and that's another great thing about the voice agents too. It's like those can also take actions. So it's like, Sure. It's cool. Like a lot of very basic AI solutions. And this is one of the ways of like, there's a sort of do it yourself, which it's like, You know, everyone has those clients that it's like, you'll talk to and they're like, well, I could just do this myself. And it's like, okay, go do it then. They're like, Oh, well. So it's like, you know, if you could, you would already do it. So I've definitely had people say, I'm just going to switch back here and stop the, screen share there. so there's a lot of like do it yourself solutions that, you know, on the outside appear to do the same thing. It's like adding a chat bot to your website, not very hard, but the thing is that it's like, They're not super useful if they can't take actions because that's like a level one basic. It's like It answers questions, but it doesn't take actions. And then you can have it take actions and then also communicate across multiple channels and to multiple agents. So that's really where it's like the benefits actually come in to having these things custom made because it's like, everyone's going to have a different CRM, everyone's going to have a different calendar, everyone's going to, you know, want it to work a little bit differently. Like for instance, say you're like. A, solar company and people want to get an idea of, you know, like, would, would my house be a good fit for solar? So you can actually have AI go figure that out for them through a chat window there, because it can use, API. So it can like call external tools, like, you know, Google solar tools and stuff like that. it can use maps to find their house and then determine if it's actually in a good spot for solar. So it's like, that's an example of like, that could be a qualifying factor. So rather than just say like. Is your house in a good position? They may be like, I don't know, but it's like, what you could do then is have it, have it be like, okay, you know, I'm, I'm interested in solar. And then it says, okay, what's your address? Then it goes and looks that up and then it says, yeah, you're actually in a good spot for it. you know, how much do you currently spend on your electric bill? Like, you know, 8 million every month. Wow, we can really help you get that down. You could have someone call them then. And then like, or someone, you could have an AI call them then. Ask a couple more follow up questions on that. So say it's like they, like, left the, you know, they just left the chat, forgot about it. You could have it call them back and be like, Hey, so I just saw that you were on our site and you'd asked a few questions. I just wanted to see if there was anything else I could help you with. Could ask a few more questions, and then they, I can tell them, Hey, it sounds like you would actually really benefit from this, and you might save thousands of dollars a month. Would you like to schedule an appointment? And if they're like, Yeah, sure, I would. Then what it can do is it can go check your calendar, Find the next available appointment and ask them if that's okay for them. If they say yes, get their name, number, email, and log that in. So that's really the benefit of like, sure, you can have a QA chat bot. That's cool. But how useful really is that? But something like that, it's like you might pay someone a full time salary to do that exact thing. And if it can be automated, I mean, you don't, doesn't need to replace that person, but that person can then be focused on more of like the high level tasks of actually like, you know, maybe answering more difficult questions or actually like helping existing customers rather than just picking up the phone to answer the same five questions every single day.
Audrey:Yeah, I think what people may not actually see right is the kind of resource allocation, right? Because this is almost like your 24 seven sales rep. And if you save your sales team to do the final closing or like connect with your audience and create a personal touch, but have someone else, or in this case, AI do the kind of grant work at the start, then it actually increases your chances of reaching out to more of a target audience. While giving your audience the information they need to make that decision, every step of the way. And I think this is also a very great use case for companies in more technical industries, where the decision making process is longer, and where you need people to fully understand, If your product is a great fit for them and, you know, handle all the different objections and with multiple touch points. So I think this is a super use case for that, Jake. one question that I had was if, if it, if someone is trying to, you know, think about a use case for their company, you know, I think they could potentially have a question of how do they start thinking about streamlining their process? Because if I run a company, I'm sure there are like hundreds of processes I can optimize. How would someone get started?
Jake George:Yeah, so I would say first look at what are tasks that you or your team actually hate doing, because there might be some things that you enjoy, you know, a lot of people complain about, they're like, I like writing, I don't like that AI does it because I enjoy doing that. It's like, okay, well, don't automate it then, you know, it has to be something that it's like, first of all, you or your team don't like doing. It's monotonous, repetitive, and it doesn't take a high level of people skills. So for example, if you're doing like high ticket closing, like I'm not going to have an AI phone agent hop on a high ticket sales call and just wing it. Like, No one's going to let a robot, like, not right now. It will happen. It will definitely happen, but it's not, you know, it's not at that level because it's like, that's a lot of like soft skills, like people skills. I would also say, don't, I wouldn't automate it. Anything that's like super sensitive, like legal documents, taxes. you don't want to go to jail because of errors that it makes. But what you want to look for is like, Maybe you start by thinking like, what would I delegate to an assistant or an intern, or what's something that it's like, you could essentially train someone who knows completely nothing. You could just show them the process in a day or two, less than a week. And then they just keep doing that process over and over and over. And it really doesn't, you know, change a whole lot. Or if it changes, there's like rules to how it might change. So I'd say. It's a low level task, you typically don't like doing it, it's repetitive, monotonous, it doesn't take a high level of people skills, and it's something that could be done by like an assistant. Yeah. so you know, like, data entry, emailing, appointment scheduling, you know, you can use it for content creation, that's definitely, you know, you have to be a little bit skilled at that on your own to be able to do that, because I know a lot of people will be like, that's another thing, like, I just asked ChatGPT to write a post, so why would I need a tool to do it? It's like, well, the post sucked, sorry to break it to ya, like, it's not a very good writer on its own, so, yeah, that's where I would say to start, you know, I guess to, to get a better idea for it, of just, I would say it's really good to take some time to, you can even ask ChatGPT this sort of stuff as well, so that's another good thing of, like, what I would do is sit down. Turn on dictation. I think on, I forget what it is on Mac, maybe like command and both command keys. On Windows, it's Windows key H. That will pop up a, speech to text thing. So as you type, it's going to like fill it in there. It's easier to get a lot more information out. So if you go to ChatGBT and you explain what your business does, what your typical day looks like, and the tasks that you're doing, and then ask it, Which of these could be automated and how and which ones would be good to automate first and it's going to give you like an itemized list of which, which ones, you know, could be done, because I guess, like, I see it different than a lot of people because it's like I've built so many that it's like I just immediately know. Like if someone tells me here, here's a task, I'll just know off the top of my head, you should or shouldn't automate that, or it's going to be easy or complex, cheap or expensive. so there's quite a range because there are higher level tasks that you can. Automate. But like I said, I'll, sometimes it comes down to, I've had people say, Oh, I talked to someone else. They said it couldn't be done. I'm like, Well, it can be done, it's just complex and expensive, and you're going to need like more of a team of agents to do this, whereas if you're focusing on like I'm, you know, only, only using one agent, then it's like, yes, one would not do it. But on, you know, there's a lot of levels of like, you know, I'm sure that there's things that I couldn't automate that probably Nvidia or Amazon could obviously, because they're on like this massive scale of like all these like brilliant AI engineers. And they're literally like developing the tech and all that sort of stuff. So it's like, you know, it really depends on like How much, you know, how much of like, how important is it to you? How much money would you save? Because that's another thing of like, is automating this task actually going to save more money than you spend on a running the bots to do it. And then be either developing it yourself or having someone else develop it. Because that's another thing is like, I'll only work on projects that will have a positive ROI. If it's something that I realized where they're like, Yeah, it takes us like, you know, the intern a couple hours a week, but to automate it is going to be like, you know, 5, 000 or something. And I'm like, you know what, it's going to take so long to recoup that money that it's just, it's not, it's not really worth it on their end. You know, we're looking for, for this to actually be a benefit. And like, it's an investment, not just like a nice thing to have. Like no one wants to just like, Oh, I have this cool AI tool that saves me one hour a week. Not really that much worth it. so especially when you're doing anything custom, because it's just like, it's, you know, more in depth than we're looking for higher level tasks.
Audrey:Yeah, definitely. So I think one takeaway is like, think about your use case for yourself, for your company, figure out if it makes sense to actually leverage AI. I always start with a problem first approach. I never just go and look at like my arsenal of tools and be like, let's use all 10. It's always starts with like a problem. They are trying to solve and then finding the right solution to solve it. and that would really help you to build a workflow that fits your team fits your business objectives It really helps you to drive the kind of conversions you're looking for So speaking of conversions and you know getting people into your funnel and system I would also like to segue into a second part of our topic jake and I have both built our personal brands on LinkedIn. I think LinkedIn is a key channel for both of us to drive traffic, right? So Jake, do you have any tips or thoughts on how people can start, you know, leveraging their personal brand for their business?
Jake George:Yeah. So, I mean, I think that LinkedIn, if you're doing any sort of like B2B, sales, like that's really the place to do it. If your, if your brand is more of like a B2C sort of thing, there, it might be more on like Facebook or Instagram. You know, I've definitely worked with companies that it's like they would benefit differently from both. Like, You know, like, so like e commerce brands, for example, it's like, you're going to do better on like TOK. But if you're specifically looking to grow a B2B business, I would say that LinkedIn is really the place for that. and I would start by like, before that's another thing about like, there are things that should and shouldn't be automated. So before. Going in and thinking, cause I see a lot of people do this. Like, oh, I'll just ask chat GBT. I'll just say, write me a post about AI and then post it. And then write me another one. And they think that they've created all this great content and they're going to grow their brand. But it's like, everyone knows it just looks low effort. It doesn't, it doesn't resonate with anyone. It's just, you're just posting words to post. And it's like, that's not. a great way to go about it. So what I would say is first start off by doing research. And this is something that I, you know, should have, it's something you learn along the way. But if I could tell my past self, as I would say, like, Sit down and really think of like, what is your actual goal for your account? Like, are you just trying to get a ton of followers if you're doing like a newsletter or something, and you're doing promotions within the newsletter of new AI tools, or, you know, for example, it could be anything, any type of newsletter, but then it's like your goal might be to get the maximum amount of followers. And post a lot of like top of funnel content that like widely appeals and will easily get you 600 likes. But if you have a very specific business, you know, for example, you're doing like AI automations or something, it's good to have different content pillars of like your top of funnel, middle of funnel, bottom of funnel. It's like you have the widely appealing stuff. You can use like personal stories, entrepreneurial mindset, and then your bottom of funnel where it's like, you're actually showcasing like case studies, what you actually do, and then driving people to like a call to action. So I think what would be the most beneficial is really sit down. start checking out a whole bunch of different accounts, because you will find other people in your niche. There's like, you'll find other people in your broad niche, and then you'll find other people that are doing pretty much the same thing as you. So it's like, kind of study them, because what you'll realize is, like, when I started, I was like, I have to have a new idea every single day, and it has to be completely unique and different than anything else I posted. And it's like, sure, I mean that, that works, it definitely did get me some followers, but what you, you really want to be like, the guy or girl or person for your exact topic there. And I think it's worth, you know, sitting down, study what works on LinkedIn, study some good copywriting skills. Find other people in your niche, have a commenting strategy as well. because all of like, it all plays a part. There's not really one thing to do. There's like 10 that you kind of have to get at least decently good at to start seeing good results.
Audrey:Definitely. I, and the funny thing was, you know, the reason why I actually approached Jake in the first place was because I saw your content and I thought that was great. So it really works, you know, having the right content actually drives you to the right leads, the right partners, and basically figuring out like who is going to come to you because of the knowledge and expertise that you share. So content strategy is key. Very important. It sets a solid foundation for everything you're trying to build. But Jake, how about someone who is trying to leverage AI in their content creation process? How would you advise them to do so?
Jake George:Yeah, I would say start off, like I said, of like, come up with your strategy. You can use AI for that. I posted some, you know, like prompts to do that. Luke Matthews is someone who has a really good, like he, you know, always puts out like good prompts for like designing your content pillars. I know that he made one based off like the Josh, Justin Wells content matrix as well. frameworks, have an idea of what you're actually going to talk about. you can use it to come up with ideas. I think that's how I would start is using it to come up with ideas and maybe help you like sort of outline a post. Another thing that's good to learn is like copywriting frameworks, like problem agitate solution or like the slay framework. or like, what is it? A I D a, you know, there, there's like, there's a whole bunch of them. I'm not even a, you know, huge expert in the, frameworks field there, but it's like, What you could do is have it come up with an idea, and then ask it, like, give me a general idea, how could I write a post about, and then insert your, your topic that you're actually going to talk about, using this framework, and use it for sort of an outline, because that will kind of get you like thinking, of like, okay, so, we have section one, section two, section three, and then just kind of fill in the blanks from there, and I would do that before I would start trying to Use AI on my own to do it, either that or use a tool that has like pre set prompts where it's like already prompt engineered to produce good results from people that already know what they're doing. because if you're just starting off, there's really no way that you're even gonna know unless you've done copywriting before, unless you've done social media before. You wouldn't really know a good post from a bad post. I mean, I didn't. I had no clue like what makes a good post. So it's something that's like, you have to spend a lot of time like researching and getting an understanding. And I would definitely like know what you're doing because AI will not just take you from zero to expert, but it will take you from good to great. So get an idea of what you're doing, what makes a good, piece of content, what a good framework is. And then I'd say, start slowly implementing it to your strategy. Start off, start off with ideas, work on your prompt engineering. Start generating some posts, but maybe don't post them. Use them as drafts, add your own personal touch. And as you get better, you'll be able to recognize. what good posts are
Audrey:definitely, you should really frame that quote, right? Hey, I wouldn't get you from, you know, zero to expert, but it will get you from good to great. Now there's a LinkedIn where we post. Yeah. So if any of you are interested, like what I mentioned, Jake and I are partners on the LinkedIn launchpad project where we combine the best of human strategy and AI execution. So. In this full three in one system, you get done for you designs, you get a virtual masterclass where you understand the human strategy that has propelled both of our growth and you get that bespoke AI tool that is trained on top performing frameworks. So you can either figure it out by yourself or join us and we can help you to get started. If you're interested, check it out. link in my comments. Now, before we end off Jake, I would love to also know what are some of your favorite AI tools. And final tips for our audience who are trying to get started.
Jake George:Yeah. So I would say some of my favorite ones, I mentioned a few of them. So, like I said, like relevance is a great one. I know it sounds a little bit cliche, but like Claude, and I know like why I say that is because some people think like, GPT, why would I use it? Whatever. But it's like. It's better in a lot of ways. I would say, you know, automation workflow, which is not exactly AI, but things like make and Zapier. let me think. What are some other ones? Oh, Sybil for, calls, like recording calls and stuff, the output on that. I've tried a ton of different, like I've used like Otter, TLDV. quite a few others. I don't even remember the names of them, but it's like I've always tried out different, call recorders and Sybil is definitely, you know, really great for sales calls, which is like a lot of what I do, in for meeting notes. It's, you know, better than all the other ones. boy, let me think. What are some other ones? Yeah, those are the ones that I can think of off the top of my, oh, perplexity, gotta use perplexity. I, man, I rarely Google anything anymore because it's like, why would I? Like I, I think I use, it's, it's one of those that like kind of goes to the back of my head because it's like, if someone asks you how many times do you use Google a day? Like, would you really know a number? No. You just like, well, whenever I am wondering something, I just Google it. But you don't think of yourself as like an active Googler. Like, I don't think of myself as like an active perplexity user, but I probably use it something like 20 times a day. If I really had to think about it, like I it's helped me learned a lot of the knowledge that I've gained over the past, since it came out a lot, a lot faster, because it's just like, you, you. It finds and parses the information for you, so you find the answers to your questions. Rather than spending 30 minutes reading to find the answer, you get the answer, and then you continue working to the next problem. And if you're just finding information faster, you're learning faster, you're getting things done faster. So, highly recommend that one, and that one's free. Just go to Perplexity. Just use it. It's, it's insane.
Audrey:They should give you an affiliate link for this book.
Jake George:Please, one dollar
Audrey:for every new user sign up. Yeah, and also like, what, what other final tips do you have for our audience? who are just starting out?
Jake George:Yeah, I would say, you know, specifically related to, business automation. If you're just starting out, take a dive into tools like Make and Zapier. You're going to get, you know, you don't need to know how to code. You're going to get a lot better idea of what the possibilities are and what can be automated. Then as for LinkedIn, if you're just starting off, you know, one of the reasons that we designed the launch pad is because of all the problems that I kind of listed out there. It's like we've identified all the parts that will hold people back when they're first starting off and unoptimized profile, not knowing about what's going on. How to create good content. So they don't know what's good and bad content, what is good content and what is bad content, and then being able to actually design it, which is why we created the tool with like more advanced prompts in it. and you know, train on viral posts. Like it knows it actually knows how to write. It's not just spitting out random tax. It's writing it in a very specific. way that we've trained it on from, you know, our combined years of learning on LinkedIn. So I would definitely say to get from, like I said, to go from, zero to expert is going to take you quite a while, but to go from zero to really good, really fast, and really get your first like 5, followers in just a few months. That's definitely where I would recommend starting.
Audrey:Thank you so much for that, Jake. And thank you for joining us. It was a pleasure to have you on the show. So guys, don't forget to subscribe to the AI Marketers Podcast. And hit the bell for more actionable AI marketing insights. We'll see you next time.
Jake George:See ya. Thanks for having me on, Audrey.