Loading summary
Transcript69 lines
- [00:01]
Michael Stelzner
I've got some exciting news. We just launched the AI ticket to Social Media Marketing World. It's like a conference within a conference. If you want to go deep in your AI learning, this is the live experience you've been waiting for. Imagine spending two days fully immersed in AI training. Grab your tickets now at social mediaexaminer.com Aicon.
- [00:28]
Kate Vandervoort
Welcome to the AI Explored podcast, helping you put AI to work. And now, here's your host, Michael Stelzner.
- [00:39]
Michael Stelzner
Hello, hello, hello. Thank you so much for joining me for the AI Explored podcast brought to you by Social Media Examiner. I'm your host, Michael Stelzner, and this is the podcast for marketers, creators and business owners who want to know how to use AI. Today I'm going to be joined by Kate Vandervoort and we're going to explore a really amazing process that will allow you to train AI to work smarter for you and your business so that you could create a whole bunch of custom GPTs and create a virtual team. Today's podcast is going to go really deep and there are a lot of little nuggets that I think you're going to absolutely love. And I just want to encourage you to pay close attention to everything that we talk about today because there's going to be a massive unlock as we approach the end of today's episode. By the way, if you're new to the show, be sure to follow us on whatever app you're listening to so you don't miss any of our future content. Let's transition over to this week's interview.
- [01:41]
Kate Vandervoort
With Kate Vandervoort, helping you simplify your AI journey. Here is this week's expert expert guide.
- [01:51]
Michael Stelzner
Today, I'm very excited to be joined by Kate Vandervoort. If you don't know who Kate is, she is an AI consultant who specializes in helping businesses implement AI. She's the founder of the AI Success Lab, a hands on coaching program for small businesses. And her podcast is the AI Grapple. Kate, welcome to the show. How are you doing today?
- [02:12]
Kate Vandervoort
I'm great, thanks, Michael. Thank you for having me.
- [02:14]
Michael Stelzner
I'm excited you're here. Today, Kate and I are going to explore how to build a custom GPT Dream team. Now, before we get into this awesome conversation, I would love to hear a little bit of your backstory. Kate, how did you get into AI?
- [02:27]
Kate Vandervoort
Sure. So I did a social work degree at university a very long time ago. We won't talk about how long ago that was, So I got the social part right. And then after 15 years in the not for profit sector, in senior leadership roles, I really got over being in a system where I couldn't affect enough change. And I always had this ability to see how to do things more efficiently, more effectively, faster, better. Some systems just don't like that kind of pace. And so I started my own business 15 years ago, in fact, fun fact, we started our companies in the same year, I think Michael and I was consulting at the time to the not for profit sector. But it was the early days of social media and it was the big black hole that no one sort of really understood how it changed marketing, how it related to your customers, your audience. And as soon as I had that experience of social media, I could see really clearly, particularly for charity or more human organizations, how they could use the technology to connect better with their audiences. And for 15 years I've been consulting and teaching and working around, growing and activating customer and cause communities, so really focused on that organic side of social media. But through that journey, I've developed a really deep passion for the intersection between technology and humanity and really understanding how we can use technology to connect with people in better ways and to be more human. And so all that time, you know, the industry went very much down the ads and funnels and you have to pay to be seen route. And I kept on my own little journey of talking about organic content and really deeply understanding your customer and speaking to them in, in a way that they felt really understood. And then I feel like the industry kind of came back around again, back to being more human. And then I was using AI before the big chat GPT moment. So platforms like Jasper AI and Copy AI and could certainly see how it impacted my business and was very excited about that technology. But it was when the ChatGPT moment happened and I saw that little cursor dancing across the screen and I really went deep for a couple of months on looking at every step of the process that we do with clients or that we'd been doing with clients, and looking at how AI could enhance or could change or amplify the impact of that. And immediately I could see how amazing AI was going to be, not just for our business, but for our clients. And that's how I've gotten into the AI journey. But I jumped in early like I did with social media and riding the wave at the moment.
- [05:36]
Michael Stelzner
Love it. So tell us a little bit about what you're doing today.
- [05:39]
Kate Vandervoort
Yeah, so a lot of what I've been doing is working with businesses or companies on the culture change that needs to happen to embed this new technology. Ironically, it feels very similar to 15 years ago, the days of the wild, wild west of social media, where companies and businesses were really struggling to understand what it meant. So I'm very much in that education, culture change place at the moment and we have a process that we run with businesses that's really about helping their staff to get past their fears. Because we find if you are sitting in fear around the technology, it's very hard to embrace the opportunity and then really doing a whole AI audit and then developing a strategy on how they can roll AI out across the business. I had a social media success academy that I closed down or sold off just before COVID I was a bit done with the whole online courses and everything that went alongside that and I really resisted starting that again. But the demand is there from my community. So I've just launched the AI Success Lab, which is a more hands on coaching place where smaller businesses can really get a handle using the same processes that we use with larger businesses, but can really get a handle on how to embed that into all aspects of their business.
- [07:10]
Michael Stelzner
Well, and Kate, I think is underselling herself because she's an active member of the AI Business Society and her technical insights go very, very deep. You blow me away with how much you've learned and how far down the rabbit hole you've gone. And I'm really excited to be able to bring you on the show today. Now, there are a lot of marketers who are listening right now and I want you to help them understand the benefit of creating these custom teams. And kind of like if they get right what we're about to talk about today, what's in it for them, what's the upside? Because, you know, this audience is deep down the rabbit hole. Some of them are just going into the hole, some of them been in the hole for a little while, like you and me, and they're, they're really, perhaps there's so many places they can focus, but this is one of the big unlocks, I think, for a lot of them. So why don't you explain what the benefit or upside is to getting this right?
- [08:08]
Kate Vandervoort
Sure. And I think there's a couple of things before we get to that that are really important. You know, for a lot of us, particularly marketers, we've been exposed to these technologies quite early. And although, I mean, for people like us, Michael, we're in this all day, every day. And sometimes I forget that there are still people that haven't yet even heard about ChatGPT or had their chat GPT moment. So I think a lot of people are feeling like they've got to rush and catch up. But realistically we're at the front of that wave and I think marketers, more than most industries or roles or positions are really at the front edge of that wave. And it's changing so quickly. You know, every presentation I do, every workshop I run, every webinar I run, I always say you have to assume that in 12 months time this is going to look completely different. And heck, some days it's, you know, 12 days time it's going to look completely different. So I think it's really important for people to understand that wherever you're at in that journey, you're still at the front of the wave. We're very much at the beginning of this AI transition if you like. But one of the things that I figured out really early on because I've been teaching the same marketing principles around who's your customer avatar and what's the customer journey and where are all your touch points and what's the content strategy that helps reach them and how do you hyper personalize content? I've been talking about that for 15 years and AI just allows us to do that in a whole lot of different ways. But a lot of people are not using AI properly yet. And so they're, and I've heard some of your other guests speak about this as well that you know, they're putting in a simple prompt, they're getting out fairly average high level content and then writing off AI as it doesn't do my job as well as I do or it doesn't give me the output that I'm looking for. And one of the concepts we talk a lot about is how do we bring the best of who we are and bring our expertise and our knowledge and then look at how AI can enhance, can help us to do more, serve more people, have a greater impact rather than how do I bring my task and just outsource it to AI and that I just want AI to do my job and I want it to do it better than me is I think a bit of a flawed concept. And a lot of people are looking at how can I just get it to save the me time rather than how can I get it to enhance what it is that I bring to the table. So having said those as a bit of a pre frame, the reason most people are getting it wrong at the moment is that they're not taking the time to align AI to themselves as a business owner to their business, to their customers, and therefore what they're getting is a fairly generic output. And so the process that we're going to talk about today is how do you really get AI lining up with you as a business owner? The way that you work, the kind of business you want to run, the structure of your business, the resources you have available, and then how do you use it to speak in really personalized ways to your customer? And once you've got that, that's where you unlock the magic and where you treat AI as this thought partner or this assistant who's going to help you be more and do more rather than doing less.
- [11:40]
Michael Stelzner
So what I'm hearing you say is whether you are a business owner or whether you work for a business that you do not own. Right. Because either case is true. This can be an augmentation or enhancement of your existing skills in such a way that you can produce better output or do things maybe that you could never imagine. Is that what I'm hearing you say?
- [12:00]
Kate Vandervoort
Absolutely.
- [12:01]
Michael Stelzner
And that's the upside. I think that what we're talking about today, right, which is like imagine now layering this dream team equation into this. Like, now, what does that really mean? Like, just define that real quickly before we go on to starting that discussion. What does that even look like to have a chat GPT? Custom GPT Dream Team.
- [12:19]
Kate Vandervoort
Yeah. So what we're talking about is like a little assistant, and we'll get into the specifics of that. But they could be a custom GPT or a project in Claude. But I find custom GPTs work best for this process that we're talking about today. And it's highly trained to deliver on the task or to speak to the audience or to help you do a particular repeatable, mundane task in your business over and over again. It's like you've got a little assistant or a buddy, and what you want to do is create a team of those. And everybody's team looks different because you've got different strengths, you've got different tasks, you've got different audiences. So everybody's team looks different, but it's like you've got this team of assistants or helpers or business buddies that are there to support you in everything that you're doing during your day. And it was one of the things. I know we're going to talk about this, but one of the things that I did really early on was I decided if I was going to embed this into my business, which was the first step for me, before letting it into my client processes, was analyzing every single task that I do every single day and looking at how AI could help me increase my productivity or lift impact in that. And I think that's the place where people want to start.
- [13:44]
Michael Stelzner
Yeah. And we're going to dig into that in just a second. But what I'm excited about, the concept of having this quote unquote custom GPT dream team is these are people that are never going to complain. They're available 24 hours a day. Right. You're in Australia, I'm in United States. Right. And sometimes the time zones don't match up, but these people are going to be available whenever you need them, of the night, when you're out and about, as long as you got Internet access. And that's kind of the big unlock is it's like quite literally going to allow you to do so much more than you can ever imagine. And when you start accepting the possibility that you could double your output or productivity or creativity, all of a sudden doors open in your mind and all of a sudden you're like, oh, that thing I always wanted to do, I could probably do that now. So that's kind of the promise of what's possible. And I just wanted to plant that seed in people's brains as we continue this discussion. So you've already kind of hinted a little bit about the first.
- [14:45]
Kate Vandervoort
If I can add in there, Michael, next year I'm going to send my AI avatar. When you send me the meeting invite for 3am to do the podcast, I'm going to send my AI avatar. But today I'm bringing me.
- [14:56]
Michael Stelzner
Okay, perfect. I love that, but I still want you. So let's talk about the first step. You've already kind of hinted a little bit like. But if we're going to buy into the concept that we're going to create this dream team of custom GPTs, you already kind of hinted that we need to audit some stuff. So like, let's elaborate on that a little bit more.
- [15:12]
Kate Vandervoort
Sure. So when we're working in a larger business or company, the first step is always to conduct an AI business audit. So we look across every task, every role, every team, every department, across the whole business and get a benchmark for the aggregated value or time that a business is spending on particular tasks. And so we can break that down for the individual, for the team, for the department, for the whole business. And so things like when you do that audit, you might discover that there's 60 hours a week happening across a team of 10 people that is spent on emails. And so once we conduct that audit, we can then go, so where are the areas where AI really has strength at the moment? Because there are some areas where it's not yet quite so strong, but where are the areas where it makes sense to implement AI in the first instance? So, from a small business perspective, I would recommend that you do what I did in my business, which is even if it's a week, but I would suggest a month. Because often we have tasks that don't just happen every week, but across the whole of your business. Just start documenting. You can have an Excel spreadsheet, you can talk to ChatGPT on your desktop, you can do it in lots of different ways, but just to start documenting. There are great apps as well that allow you to track your time, just to start documenting. What are the repeatable tasks? What are the things that you're doing every day, every week, every month in your business so that you have got a benchmark? Because as marketers, we like to measure our return on investment. And if you're going to go into implementing AI across your business, you want to understand where you're starting from, because then in a couple of months time, you can really see start to see those productivity gains. So in our coaching program, we actually have a custom GPT that people can use to have a whole conversation. But if you have an Excel spreadsheet where you've tracked your time, you can upload that into Claude, you can upload that into chatgpt, into Copilot, into Gemini, whichever platform you're relying on the most. And you can start to have a conversation around, this is what my week looks like. Now, what can you do to help me streamline some of those tasks? Which areas should I focus on first? So you can conduct that audit yourself and it doesn't need to be complicated.
- [17:44]
Michael Stelzner
I love that, by the way. And you could probably even take a screenshot of it, you know, because these things are really, these vision models are crazy. So basically what I'm hearing you say is, hey, figure out a way to document in some sort of a spreadsheet, like the kinds of tasks you're doing and the amount of time you're spending doing it, presumably. Right. And then you can identify the obvious ones that will be candidates potentially for assistance with AI. We talked about an example about a nonprofit. Can you share that example?
- [18:12]
Kate Vandervoort
Yeah. So we've got quite a large nonprofit that we are working with on an ongoing basis. And one of the tasks, and this is just one of the tasks but one of the tasks that they were doing, they run a residential facility, and one of the tasks they were doing was inputting handwritten feedback forms from their participants every week. So there'd be 40 or 50 handwritten feedback forms that the administration person needed to hand type was still hand typing, and it had to be handwritten by the participants. There wasn't an option to have it be an online form. And there were some reasons for that, but they completed the form. What we ended up doing. So that was taking them six hours a week. And what we ended up doing with them was the admin assistant can now scan 50 pages into the photocopier. So it's a PDF file. You could also take photos of handwritten notes. We've done a lot of that. You know, you do a brainstorming session, you take a photo, put it into ChatGPT, it will read it and transcribe that for you. But that was a job that then took one or two minutes to get all of that transcribed into an Excel spreadsheet. And then we set up a custom GPT that allowed some analysis to happen, and we gave it a formula of what we wanted the sentiment analysis to look like. It would pick out the top three testimonials, it would pick out any negative sentiment that was in those feedback forms, automatically write an email that can then be sent to the general manager for review before it was sent to the person providing the feedback. So Something that took six hours was reduced to 20 minutes.
- [19:59]
Michael Stelzner
I love that.
- [20:00]
Kate Vandervoort
And when you look at those repeatable tasks that happen every day, every week, and you just start to shave off that time, the aggregated amount of time saved is enormous across an organization.
- [20:13]
Michael Stelzner
I love it. Okay, so step one is to document audit, if you will, all the things that you or someone who works for you or multiple people that work for you are doing to try to identify some areas, and then ultimately ask AI to assist you, in which one of these areas do you feel like you can help? I love that Tip.
- [20:30]
Kate Vandervoort
Yeah. And if you have a team upload all five of the spreadsheets, all 10 of the spreadsheets, and say, tell me across our organization, and obviously you want to make sure that the columns are consistent, but across our organization, how much time are we spending on email? Okay. Chatty. I like to call chatty. Chatty. Okay, Chatty. You know, if we're spending.
- [20:51]
Michael Stelzner
That's your name for ChatGPT?
- [20:52]
Kate Vandervoort
You asked?
- [20:52]
Michael Stelzner
Yeah, that's your name for ChatGPT. Keep going.
- [20:54]
Kate Vandervoort
Yep, that's my name for Chatgpt. So you know, hey, Chatty. So we're spending 60 hours a week across our team of 10 on responding to and writing emails. How can you help us do that better? How can we save time on our emails? And you can then just have a conversation And Chatty's very generous in helping you out and figuring out some of those productivity gains.
- [21:19]
Michael Stelzner
Many of the top experts you've heard on this show will be speaking at Social media Marketing World 2025. And with your AI ticket, you can attend at a very economical price. You'll discover practical AI workflows and advanced AI automations that will increase your productivity and save you time. Imagine getting live and in person training from Matt Wolf, Chris Penn, Jeff J. Hunter, Rachel Woods, Molly Mahoney, Brian Piper, Jeff C and many others. Isn't it time to enhance your career by fully embracing AI? Grab your tickets now at socialmediaexaminer.com Aicon I love it. Okay, so what's next? Once we've got this information and guidance from AI, what's the next part of the process to ultimately get into this dream team?
- [22:11]
Kate Vandervoort
Yeah, so then we're really looking at how do we align AI to us as a business owner or an individual or a professional to our business, the type of business, the industry, and then to our customer. So our process is that we create 10 foundation documents that are going to be used to train whichever AI it is that you're using. But let's stay focused on the custom GPT Dream team. So that first segment is, by the.
- [22:44]
Michael Stelzner
Way, what is a foundation document?
- [22:45]
Kate Vandervoort
Yeah, so a foundation document is like a strategy, a script, an example, a profile. And these are documents that we can then upload to AI so that ChatGPT references those documents and the output is completely tailored to the context that you've provided. ChatGPT.
- [23:07]
Michael Stelzner
I love that. Yeah.
- [23:08]
Kate Vandervoort
And so this process, we start with the professional me document. And when I did this the first time, it was mind blowing how just that one step changed the output of AI. So in this document you want to put in everything about you. Your background, your bio, your experience. I know in America, university and scores and all of that are a lot more important than they are here in Australia. There's a lot more focus on that. You can put that into your professional document. You can put all the information from your LinkedIn profile, you can put in your personality profile. So Myers Briggs Disc strength finders, whatever your profiles are, you put that in. You know, we've done some amazing things with people who have ADHD or other Learning difficulties. You can put in your strengths, you can put in your weaknesses. If you don't know all of this, you can work with ChatGPT to discover a lot of that. And you know, there's this viral prompt that's going around at the moment of asking ChatGPT, what's something that you know about me that I might not know about myself? And I often chuckle that I'm sure AI sitting there going, geez, these humans, they're so needy. They want constant validation. But you want to create this first document, which is all about you as the business owner, because then everything, or the marketer or the professional or the team leader, because then everything that AI gives you after that is aligned to you. So one of the things that I'll be the first to admit is that people management and running big teams is not my favorite thing to do do. It's not my strength, not my favorite thing to do. So when we go through this process of building a business and understanding how best to run that business, knowing that is going to change how ChatGPT suggests that I build my business or suggests.
- [25:11]
Michael Stelzner
That I work real quick on this foundation document. So we're building this document that describes, in this case, us, what do we do with this information. How does that work its way into AI, if you will?
- [25:23]
Kate Vandervoort
Well, there are 10 documents, so we've just talked about number one. And we're obviously not going to go into detail about all of them. But the way these documents work is in this process. We go through an iterative stacking process. So I'll just give an example. When you go to the second document, which is about the business, the structure of the business, the industry that you work in, all the information about how your business operates, the resources that are available, the staff you might have. What we want to do is also draw in the professional document to that, so that as we're building this second document, we're bringing in the first document. And as we go through this process, we're drawing in the previous document so that it flows all the way through. And what you're building, the end outcome is highly aligned AI that is speaking directly to you as the business owner, to your type of business, and directly to your target audience in ways that you just can't do with whatever however fancy the prompts are that you're using in AI, because those documents we then upload into the custom GPTs at the end. And so it's actually referencing these documents that have gone really deep. Now, that's the process that we Run, but everyone listening can do that themselves. So just put together a document about who you are, how you like to work, your strengths, your weaknesses. You know, AI is fabulous at filling the gaps around our weaknesses, and so you can create that document yourself. It doesn't need to be a fancy process.
- [27:04]
Michael Stelzner
Part of the reason I think this is interesting and that it works is having interviewed quite a few folks and listened to a bazillion podcasts. AI is really good with mimicry. It's really good at modeling data that you give it. And I would imagine the reason why this is so important as you give it information about yourself, information about your business, information about how you do your marketing. You can have a whole one on your core audience and your products and everything. I can imagine this is going to allow, first of all, its database to narrow the focus, right, of what it does. Instead of focusing on the whole universe, it's focusing on. On this. But I would imagine it's also looking at the way that this stuff is crafted, and it's going to model some of that language. So you probably want to write it in a way that you want it to model. Am I right on that or am I totally off on that?
- [27:52]
Kate Vandervoort
That's not so much the purpose of these documents. And maybe I can just map out the 10 documents so that you kind of get that bigger picture. So it's kind of in three sections. So the first section is the professional me, the business basics, and the market analysis. And so what we're doing there is going deep on who am I, what's my business, and what's the market that it sits in. So we do market trends, we do a SWOT analysis. And these are all models that you can just ask ChatGPT to run for you. It will look for your competitors. You can put your own competitors in, you can do an analysis on. Here are my top five competitors. If you map the market trends to your competitors and their offering, where are the opportunities, where are the gaps, what's not being filled at the moment? And it was interesting when I did this process in my own business, one of the really, given who I am, my history, my background, my experience, One of the conversations I had with ChatGPT that really switched this on for me was how important the need for human centered, ethical, responsible implementation of AI was. And that was uniquely in the conversation, based on my history, my background, my focus, my passions, my interests. So that's that first section. The second section is we look at what's the portfolio of services, of products, really mapping Those to the market trends and then who's your target audience? Now that's a process that marketers are very familiar with. But some of the things that you can do with your Personas, your target audience, you know, we've done things like, and this used to be weeks of work in our business. And in fact we had a, an app on our website that was an avatar generator that went through, it was about a 15 minute exercise went through asking a whole lot of questions that created an ideal customer avatar for a business. Once I realized that ChatGPT could do that in moments, that was a big for me. But one of the other things that we've done over the years is focus groups, surveys, really understanding the pain points, the challenges of our target audience. Once we developed the target audience using ChatGPT, we then got ChatGPT to act as a focus group researcher and conduct a focus group with our ideal customers. And it actually mimicked. It was literally that moment for me just exploded everything because I just went, that is weeks or months worth of work. Because not only did it mimic Mary, Craig and Jo having a conversation about our products and services, talking about their role, their position, their pain points, how our products and services could actually help them address their needs, we then could take the output of that and use that in the rest of the creation of these documents. So we then have a focus group document that has got feedback and the language that our customers use. And I can't overstate the amount of time it used to take us to do that in our business.
- [31:24]
Michael Stelzner
Well, and just for people that want to go down this rabbit hole, episode number 11 with Brian Piper building custom GPTs, personalizing an AI assistant. He works at a university and he created an avatar of students that were thinking about going to the university. And the whole marketing department tests all their ideas up against that avatar to see whether it's going to work or not. So that's a fascinating unlock. So, okay, so I think you mentioned six out of the 10 so far.
- [31:50]
Kate Vandervoort
And if I can add another fun example of that, we've done things, particularly in the not for profit sector where we've created an AI board member avatar and they run board papers and ideas through this. Highly trained, highly aligned. So before they ever send the board reports off to the actual board, they run it past this AI. I mean there's so many ways that you can use this, this, but that was one of the big unlocks for me, was actually getting it to mimic your Personas. And you can see, I'm sure for Those marketers that are listening where this is going around, how you can then align everything else to that. So that's that middle part is really around the products, the services. But this iterative stacking means that when we're working on the Personas, we want to feed in the market trends, the competitors, the products and services. So you're actually layer in all of the documents as you go. And that's when it starts to get really exciting because the next segment is around a business plan and a financial plan. And so instead of just doing your business plan for 20, 25 or whichever year you're in, if you're listening to this in the future, you can create your business plan and your KPIs and your targets, all based on market trends, aligned to your target audience, aligned to your business model, and the industry that you're in aligned to you as a business owner, if you're a solopreneur. So everything that we do from here on in is referencing those previous documents. So that's kind of the foundation work that most marketers do. And then we get into the last part, which is mapping the customer journey. And again, that would take us weeks sometimes to do with larger clients mapping out a customer journey, looking at all the touch points. And then we do the content strategy. So what type of content? And you know what's really important? If you've identified in document one that you're never going to do dancing TikTok videos, well, that's going to carry all the way through and you won't have chatgpt suggesting that you do dance dancing TikTok videos at the end because it already knows how you want to approach your business.
- [34:12]
Michael Stelzner
I love it. Okay, so what I'm hearing you say is you can create lots of these foundation documents, which could be in a Google Doc, you could create PDFs out of them, or I would imagine you could even copy and paste some of this stuff, depending on what model you're using. Right. Like Claude, for example, allows you to copy and paste or upload PDFs. In the case of ChatGPT, we're going to get into this a little bit. But having these documents is the key to everything. Right? Going to take a little bit of time to make these documents, but once these documents are made, they can all kind of interactively work together. So that's, I think my next question is like, now let's, let's assume people are tracking with this and they've created a few of these, maybe to begin with, you know, how do they ultimately use these documents to ultimately create this dream.
- [34:58]
Kate Vandervoort
So just a couple of things on what you've just said. You definitely want to save them either as Google Docs or you can save them as text files. But we have found that text files and Google Docs or Word Docs work better than PDFs. That's just a little, a little trick because you're not cutting and pasting this information ever into places. We're going to use these foundation documents in the backend of a custom GPT so that it's referencing these documents. And the beauty about that is that you can switch these documents out over time. So as your business strategy changes or you introduce a new product or you have a new target audience, you can then switch those pages out, all those documents out in the back end of the GPTs. So that customer focus group is one example. But we during that process usually run a couple of different activities, I guess that are strength tests. So you could, if we take it from the viewpoint of a marketer who's listening to this podcast, you could go to ChatGPT and say, I want you to act as a McKinsey level consultant and tell me which models or frameworks you use to develop a marketing strategy. And ChatGPT is going to give you those models and you can then feed these documents into those models. So you can do a blue sky thinking process. You can run a business canvas. You can do all sorts of things where you're testing these. So for example, I do a blue sky thinking process using my foundation documents quite often. And it'll suggest five things like create a hologram of yourself to turn up to workshop. And then I say, you know what, those three out of those five are not going to work for my business. So what else could you suggest? And you can say, be quirky, think out of the box. What else can you suggest to me?
- [37:05]
Michael Stelzner
Okay, so we've got a document we can export out of Microsoft Word or out of Google Docs as a text file or as a Word Doc. Right. So let's just say we're exporting it as a Word doc. Help me understand, like we've got all these different documents and maybe this is something because I've never made a custom GPT. I've only used Claude projects. So help everybody understand, like what do they do with these documents? Are they just uploading them into a thread that they're having with ChatGPT or is there something more going on here technically behind the scenes?
- [37:34]
Kate Vandervoort
Sure. So custom GPTs are these little assistants that are trained to do a specific task, are trained on a particular product, or maybe to speak to a particular audience. What we do is when we create these, and they're very simple to create. Anyone that's out there saying that custom GPTs are complicated, they're very simple to create, and I'll give you a couple of tips around that shortly. But when you create the custom GPTs, there's a section in there called knowledge files. In those knowledge files, that's where we're going to upload these documents and then in the instructions. So let's say. I'll just use an example. Let's say you got a Facebook group custom GPT. Its sole purpose is to create content for your Facebook group. Now, if you just ask ChatGPT, create me some posts for my Facebook group about AI, it's going to give you some fairly generic posts. If you upload these foundation documents that are about your target audience, the market trends, your content strategy, and potentially what your goals are for your business that year, what the custom GPT can do is map your audience against the customer journey in the context of your business plan and your KPIs for that group. And then you've got a custom GPT that when you go into ChatGPT, instead of just having a normal chat, you go into your custom GPT and then when you press the button that says create me five Facebook posts, it's drawing all of that information in. And what it produces is so far removed from just an average ChatGPT chat that that's where you really start to see alignment. And I think that's the word that everyone needs to keep thinking about is it's how are we aligning AI or aligning ChatGPT to us as a business owner, our business, and our audience.
- [39:46]
Michael Stelzner
Okay, So I think the dots have started connecting in the minds of lots of people. And what I'm hearing you say is you go and you create a custom GPT, it's going to ask you to give it source material, and you're going to upload pretty much all of these documents, right? These foundation documents or whatever ones that are most relevant, right? And then then basically what happens next, like, help us understand, like, what other, you know, we've got these documents in this custom GPT and you advise you have a bunch of them. It sounds like, like one, for example, just for Facebook group posts. So, like, what else do we need to understand about these custom GPTs?
- [40:20]
Kate Vandervoort
Well, in terms of determining which ones you need, that's where I would be Uploading that AI audit from the beginning about all your tasks. And I would upload your business document and maybe your products and services. And you would ask ChatGPT, if I was to create a team of 5, 10, whatever number you want, you don't want hundreds of them. But if I was to create a team of custom GPTs, given everything you know about me, what GPTs would you suggest that I create? So that's the first thing. And ChatGPT will go, you know what, Kate, you run a Facebook group and how you post in your Facebook group is different to how you post on LinkedIn. So I think we're going to give you two separate ones here.
- [41:05]
Michael Stelzner
Okay.
- [41:06]
Kate Vandervoort
And I think one of the things that's important is, you know, in our coaching program, we have micro businesses sometimes that are literally one person. They've got nothing else going that is in their business other than themselves. They might not want a Facebook GPT, a, a LinkedIn GPT, a blog writing GPT, a lead magnet GPT. They might want a marketing GPT because they can then.
- [41:34]
Michael Stelzner
Or a content marketing GPT.
- [41:35]
Kate Vandervoort
That's right, that's right. So they might chunk up, if you like, and look at kind of a more of an overarching one. Larger businesses. Absolutely. You want more tailored GPTs. So then what we do is we go into ChatGPT. You go to Explore GPTs, you click on Create and there's two tabs there. One is called Create and one is Configure. Now, Create, you can just have a conversation with ChatGPT and it will create it for you. So maybe the first time you do it, go and just have a chat with Chatty and tell it exactly what you want to do and it will create it for you. Having said that, the process we've just gone through with the knowledge documents, you're way more advanced now than just the Create tab. So you just go straight to the Configure tab. You can upload an image that represents the custom GPT. You give it a title, you give it a description, then you've got the instructions. So there's a couple of key elements here. The instructions are critical. You want them to be simple, you don't want them to contradict each other. We have a framework that we usually use around what's the role or the position of the custom GPT. So you're an advanced marketer, you're a focus group facilitator, so you give it the role. Or you're a Facebook marketer, you give it the role. Then you give it the task, so what's its purpose? Why does it exist? And then the instruction of how it behaves. So the tone that you might want to use when I ask for you to create a marketing strategy, I want you to outline the KPIs, I want you to put them in a spreadsheet so you're actually giving the instructions of how you want it to behave. And I'll come back to the custom instructions in a minute. Then the next section is the knowledge files, and that's where we upload any documents that might be relevant to the task. The product that GPT is representing.
- [43:43]
Michael Stelzner
I love it. So you said you were going to come back to something. What was it?
- [43:47]
Kate Vandervoort
Yes. So one of my favorite ways, and I do this a bit differently now because I've been doing it for a long time, but one of my favorite ways to do this early on, and this is a major unlock for marketers in particular, if you're having an amazing conversation with ChatGPT, and I'm sure we can all remember those moments where it's like, oh, we've nailed it now we've got got the formula, or you've really done exactly what I've asked you to do. At that point, you can literally, in that same chat, say, hey, chatty, I want to capture everything that we've just done here, and I want you to write me a set of custom instructions for a custom GPT. And what ChatGPT is going to do is capture what you've talked about in a set of instructions that you can literally cut and paste across into your custom GPT.
- [44:41]
Michael Stelzner
I love that and I've tried that. And that's amazing because sometimes you're like, oh, I got this thread and it's beautiful, right? And. But eventually they don't get very smart because you lose, like, they lose some of their memory. Right? And that's where. Is there a specific way we need to word that to get what we want out of it? Or is it. Can you just be very simple and just say, hey, I want to create a custom GPT. Would you write me an instruction set? Is it that simple?
- [45:02]
Kate Vandervoort
Yeah, I mean, we have a bit more of a detailed prompt because we have a prompt structure that we use, but you could apply any of the prompt structures that all of your guests. You know, there's so many prompt structure frameworks at the moment, we've created a really simple one, which is the SNAP formula. So set the role, name the task, add the context, and then your preferred format. So it's just four quick steps that the instructions could follow. Yeah, but ChatGPT knows what to do, though, in terms of creating instructions.
- [45:35]
Michael Stelzner
Well, and I would imagine it might even know what this snap method is, even if you created it so someone could ask it, hey, do you know the snap method of creating a prompt? And if it does, tell me what it is. And then boom, all of a sudden it's like, crazy. So, all right, so what I love about this is we've done a lot of groundwork and we've gotten to something really exciting now, which is we can basically create as many of these custom GPTs as we want. They can be private or public, is my understanding, based on everything I've heard. If you have teams accounts, you can share them between different staff members, which is even more exc. And we can create all sorts of different, quote, unquote, virtual team members. Now that will allow us all of a sudden to do very specific tasks. Am I hearing that right?
- [46:18]
Kate Vandervoort
Yeah, absolutely. And I'll give everybody a little. The first one you want to create is an advisory board GPT for your business. And I love this process. So I had a conversation with ChatGPT and I said, right, these are the main functions of my business. I'm a consultant, I'm a course creator, I'm a content creator. These are the things that I need to do in my business. Who. The leading experts in the world on these areas. And Chatty will just lay it out for you, you know, course creation, Amy Porterfield, Pat Flynn. It'll outline all of the experts that are in that space. And then what I did was created an advisory board.
- [47:01]
Michael Stelzner
Oh, like a virtual Amy Porterfield.
- [47:04]
Kate Vandervoort
I might have a Michael Stelzner in my advisory board social media perspective, you know, Ryan Levesque on funnels and Feifei Li from Stanford on human centered AI. So I create an advisory board that is a custom GPT. And then every piece of content that I create, I can run it through the advisory board. Every idea I have for my business, I can run it through the advisory board. And Michael might say, kate, your social media strategy here is not strong enough. Or, you know, Ryan Levesque might say, why don't you develop. Develop a quiz funnel? It would work really well for this one of your customers. And so if you've actually brought in those foundation documents, you've then got the world expert leaders providing advice based on just your context.
- [47:52]
Michael Stelzner
I love that. It's like having a mastermind without having to pay for access to all these people, Right?
- [47:57]
Kate Vandervoort
Totally. And just so you know, I have paid a lot of money to all of those people in the past. I'm not just ripping them off.
- [48:04]
Michael Stelzner
No, I love it. It. Kate, this has been absolutely fascinating. If people want to discover more about you and your business, first, what's the preferred social platform for them to connect with you? And secondly, if they want to work with you, where do you want to send them?
- [48:20]
Kate Vandervoort
Yeah, look, I'm mostly on LinkedIn now. I do have a Facebook group called the AI Success Lab, but you can find me on LinkedIn as Kate Vandervoort. And we do have a pretty cool AI success strategy process that is totally tailored to you, your role and your industry. So that sits@aisccess strategy.com you can go and find that there.
- [48:46]
Michael Stelzner
Kate, thank you so much for sharing your insights with us today.
- [48:48]
Kate Vandervoort
Thanks, Michael. It's been great being here.
- [48:51]
Michael Stelzner
Hey, if there's anything we talked about today and you missed it, I got you. We got all the notes over@social socialmediaexaminer.com A31. And be sure to follow this show. We've got a ton of great content coming your way and if you've been a regular listener, I would love it if you would consider giving us a review on whatever platform you're listening on. And we could also use your help getting the word out about this show. So share the show with your favorite friends. You can tag me on Facebook, at Stelzner, on LinkedIn, Telzner, and on X at Mike Underscore. And we've got some other shows also. If you love this show, you might also love the Social Media Marketing podcast and the Social Media Marketing talk show. This brings us to the end of the AI Explored podcast. I'm your host, Michael Stelzner. I'll be back with you next week. I hope you make the best out of your day and may AI help you become more successful.
- [49:45]
Kate Vandervoort
The AI Explored Podcast cast is a production of Social Media Examiner.
- [49:53]
Michael Stelzner
Don't forget to get your AI Ticket to Social Media Marketing World 2025. Become an AI Enhanced Marketer. Grab your tickets now at Social Media Examiner.com AICON.