Fred Faulkner (00:02)
So as soon as custom GPTs came out and I was already paying for OpenAI, that was like, great. Now let's go dive in. How can we start to build things and learn how to create context and background and other information and really get us to use and be more efficient in our jobs?
So I was all in as soon as those things kind of came out and experimenting with which versions are best, right? There's chat GPT, there's gems from Google, there's projects from Claude and Anthropic. So what works best for what things you just have to experiment, build your use cases and experiment. So I was in as soon as they came out because I had to just figure out where this is going to help my team become better.
Ben Ard (01:02)
Welcome back to another episode of Content Amplified. Today I'm joined by Fred. Fred, welcome to the show.
Fred Faulkner (01:07)
Hi, Ben. Thanks for having me on. Appreciate it.
Ben Ard (01:09)
You bet. Fred, I'm excited. We're going to time dive into a really fun subject today. Something that I think is kind of like the next level of AI that a lot of people need to embrace. But before we get there, Fred, let's get to know you. Let's get to know your background, work history, and let the audience kind of understand who you are.
Fred Faulkner (01:27)
Sure. Sounds good.
So my name is Fred Faulkner. I am the SVP of marketing here at McFadden Digital. McFadden Digital is a global commerce system integrator where we implement commerce technologies and help our clients get ready for the future of the AI world. Specifically, we focus on B2B manufacturers and distributors. based here in the US but have global presence and really happy to be here on part of the show.
Ben Ard (01:47)
I'm excited. This is going to be a ton of fun. And with your experience, you get to see a lot of the AI stuff that's cutting and bleeding edge. I think this is going to be fun. What we're going to really dive in today, we kind of came up with a pseudo episode title of building your AI bench, how custom GPTs can help you create without losing your voice. This is an area where I am fascinated because I have so many times where I love my custom GPTs and I hate them.
Fred Faulkner (02:12)
Yeah.
Ben Ard (02:12)
So Fred,
I'm honestly like excited to dive into the conversation, but what led you, let's start here to like first start building custom GPTs. Like what got you to start thinking, okay, this is worth my time and effort.
Fred Faulkner (02:28)
Sure. think to start to get to that answer, I'm to do a quick backstory. So I grew up in the first age of the internet and I really saw the potential where the internet was going. I was back in college where I was a marketing degree and really wanted to, I kind of sit at the intersection of marketing and technology. So I've always really been fascinated by technology from the get-go, building websites, doing email marketing, SEO, kind of all those new content, all the kind of things going, you know, back in the late nineties and the two thousands, we had the big bust. Things kind of of came back.
But I've always been fascinated with technology. Really kind of gone back to even like my family. had my first 2X CD-ROM drive. I've had computers in my life for a long time. So when it came to November 30th, 2022, when ChatGPT first kind of like came on the scene, I've been aware of AI and ML and AI in general and machine learning and all the different things, but I've never been tech savvy enough to really be able to execute on that type of stuff. But seeing the ability to interact with AI in a
conversational and kind of new interface. Like I knew we were entering a new chapter of what potentially could be marketing and business in general. I mean, a lot of other people did too, don't get me wrong. We've seen obviously since then, last three years have been everything's AI. But I did kind of see like as a marketer, saw like there's difference. This is different than what we've seen before. A lot of potential. So I think I've been really trying to create what would eventually would be a custom GPT for a long time. I have these really long prompts. I really have a separate document that every time I'd
Want to do something? I load all this other context in for the context window to try and get it to do the things that I wanted it to do. And then custom GPTs came out and that's when I knew like, all right, this clearly now we're getting this vision. I think everyone's seen Iron Man. They want their own version of Jarvis. What can you do with this type of stuff? And this was like, I have the things that go through your head. So custom GPTs for me was this ability to start to really look at what I do as a marketer, look at my team, which is really small, only about three people.
And understand like, can we leverage these tools, like a custom GPT to help us do our jobs better, more efficiently? Because I am very much a, AI is a counterpart. It's the Robin, the Batman, where the human is the Batman and Robin is your sidekick. It's never to replace what we do. It is here to augment and advance our ability to do things faster, more efficiently and effectively. So, and knowing that every AI is that what you're interacting with today is the dumbest it'll ever be.
knowing it's only going to get better, then you have to just start experimenting.
So as soon as custom GPTs came out and I was already paying for OpenAI, that was like, great. Now let's go dive in. How can we start to build things and learn how to create context and background and other information and really get us to use and be more efficient in our jobs? And as a person in general, that's passionate about AI as well as a marketer and a leader of a team, there was just...
kind of had to fall into it, you know, and just kind of go. So I was all in as soon as those things kind of came out and experimenting with which versions are best, right? There's chat GPT, there's gems from Google, there's projects from Claude and Anthropic. So what works best for what things you just have to experiment, build your use cases and experiment. So I was in as soon as they came out because I had to just figure out where this is going to help my team become better.
Ben Ard (05:30)
I love it. Very cool. So for everyone listening that isn't super aware of what a custom GPT is, I think this will kind of go into our next question as well. Why would I use a custom GPT? guess what is a custom GPT just to catch anyone up who's not aware of that. And how do you personally train your GPTs in a right way to get your style, your voice, to do what you want it to do. Everyone's got different techniques and strategies.
Some of them are better than others. Kind of curious of when are you creating GPTs and how are you ultimately training them?
Fred Faulkner (06:03)
Yeah, I do it in two different ways. And I have two different use cases on this. But let's talk about what a GPT is to start off with. So a GPT, custom GPT, is the idea of taking all the power of what OpenAI or any platform, AI platform, you have. They have their different versions. taking the power of what those platforms are and all the learnings all the elements of what has happened from all the elements of what they've learned and trained. And now put it in a sliver that's just specific to either a use case or something that you want it to do.
So you have all the power of like data transformation, creativity, images. You have all these kind of check boxes. can enable all the different standard features, but you basically give it a task or you can go broader as narrow as you want. I suggest going narrow, kind of find a use case so you can really make it more specific versus like a broader system. I have variations of both. And it allows you to create custom instructions, give it knowledge and background. Now you can connect it to actual your own systems, whether it be your productivity tools like.
Gmail or Outlook or Microsoft tools or your workflow tools like Asana and other stuff. All these things are now like an ecosystem is building up. the point of it, think, is at the end of the day is to give you the ability to give it a narrow scope of work of what you want to do, but give it context of how you want it to function. So custom instructions, things like, I am a very conversational person. So I like my writing. I have a writing assistant. I make my conversation, my writing, ⁓ when we work on writing things together.
Make it conversational, because I'm a conversational writer. I like to use dashes. my gosh. You know, that's a big kind of thing, you know, right now. was like, you can tell it's AI because it uses dashes. Well, I actually write with dashes, like just kind of the things I do. I, well, I have to go back and like rethink it now because everyone thinks I'm, you know, there's there's a continuous and now a stigma around it, but you can give it custom instructions. So how you want to respond to yourself, where it wants to, how you want it to behave, like give it a.
Ben Ard (07:30)
I try to train that out of my stuff and I can't get it to happen. ⁓
Fred Faulkner (07:44)
⁓ you know, any kind of nuance of how you want it to act. you are a persona. you know, one of the things that are big in AI is like, you want to give a prompt, a persona so you can give your custom GPT a persona. You are an expert level writer in this industry doing this type of stuff. Right. So that gives it the context of how it wants to communicate with you and what it's, what it's kind of. Field of view is I would say, then you have the ability to give it actual content to refer to. So I have not only custom instructions that I.
outline my tone and voice and other attributes I want to act on is then I also have given it all my podcast transcripts, all my blog post transcripts. I'm going talk about my personal assistant and then I'll talk about it in business roles here in a second. But I've given it a style guide, legit things that like how you want it to reference. So these are not only tools that I could reference saying, you've written about this before. So I'm going to highlight this because I know that you've already written, but it also gives you the idea of getting more
perspective of your tone of voice because you've written these things out. It's actually reading your style and tone. So it has a reference point. And I think good prompts give you reference points of what the output should look like. So this knowledge base helps it understand what the output should kind of fundamentally look like. So you're giving us the best of what you've created before. So those are two ways that I do it. The third way that I help my GPT understand and how to work with me is, and especially get the right tone of voice out, is I actually talk to it. So not write.
I actually talk. And so when I'm working with my custom GPT, Aria is my writing assistant. So I talk with Aria and I have an idea, whether it's, you you're in the shower, you're on a dog walk, you're I really, you read an article, like, I want to have a point of view on that. I will fire up Aria, my custom GPT on my phone, and then I'll actually go into advanced voice mode. And I will just start brain dumping. And I find that that is like one of the easiest ways for not only for Aria and my custom GPT to understand.
how I talk, what my thinking process is and how I go about doing it. But it also then has the context of like, again, tone and voice and style and the conversational component. And so by brain dumping, I also then give it the context of like, all right, now reformat this, my words, my tone, my text, my voice, re help me reformat this into an idea, a blog post, a podcast episode, potentially a white paper. Like, so that's how I use the custom GPT.
tell me even some of the stuff that I do. I have a strategic thing, like a strategic co-pilot that I think about strategy. have like a chief of staff that helps me with some of my job. So like for those, when my work related ones, I've given it my job description. I've given it my goals, my KPIs I need to hit. I've given it all this other context of what I need to do as a person, a SVP of marketing. And then context about my work and stuff. Again, I'm in a world where we're siloed off. We're not trained in the model. Be very careful about that. as you know, just the default, never train the model. So whatever, turn that off always.
But I give that to help me help it understand how I need to do my job better. So those are the kinds of things that I use and to give it context, but also talking to it, I will say is one of the biggest things that I would say is help me not just get my thoughts out on paper because thinking and then writing, you're always, you're always self editing when you write, when you type something, you're always self editing. Like, I don't want to write like that. I want to write like this, talking to it just brain dumping stuff out by far is the best part of my workflow right now when it comes to working with my assistants.
Ben Ard (10:49)
I love that.
Well, it's kind of an interesting thing. So I love that you're talking about it because the talking to your GPT, you're right. Like it gets your ton of voice. It really captures how you think, how you speak, how do you say things? Because no matter how good of a writer we are, we speak and write differently. And if you want it to feel and be authentic as possible, the ability to just have.
Basically a ghostwriter understanding your tone of voice, how you say things. You're, you know, the vernacular that you typically, you know, use all that kind of fun stuff. That's so cool that it's able to do that. But what's interesting for a lot of people that aren't using AI a ton, this feels awkward. Like it really does. You feel like you're talking to a computer. How have you gotten over that? Because I've seen it time and time again. I'm curious what you think here.
Fred Faulkner (11:30)
It does. It does.
⁓
Well, first of all, kudos to like the actual LLM platforms. They've addressed this. At first, it is awkward. I will say advanced voice mode, specifically in OpenAI chat GPT, is uncanny unreal, realistic. Like it is. It has gotten so good where it actually inserts the ums and the ahs and the pauses and you can get very quickly forget you're talking to an AI assistant.
Ben Ard (11:57)
Yeah.
Fred Faulkner (12:03)
Before you just can't like the way that if you built your custom GPT and you kind of tell how to respond to you, like, please refer to me as Fred, please, you know, you have a conversational tone. Please challenge my thinking. Like you give it all those kind of context. Just don't please me and say yes to everything. Like you got to give it the parameters, give it the guidelines. But I was doing something where the other day I was a couple of weeks ago, I was working on like my 2026 strategy and it was late and it was at night. I already had like a full day at work and like I knew I had to get like a draft done or just get some thoughts on paper. And so I literally was just.
talking to my GPT, to my strategy one. My wife came in after, she's like, were you talking to, who are you talking to? Because it's like a British woman voice, you like you could customize, right? That's like eight to 10 voices. Like you can give it this persona of like it's a real person. And like, I was talking to you and the name of my GPT. And she's like, you've given it a like, yes, I actually have given them personas. I have Friday. I have Edith. have, you know, I have all these different ones. I've just named because they have different functions for me. But.
Ben Ard (12:38)
the
Yeah.
Fred Faulkner (12:58)
And you, these advanced voice modes are just getting way better. It has a lot of byproduct and downsides outside of this conversation of what is actually doing for a lot of humans and what they're doing when they interact with them. it is, Dave, I mean, if you think about where conversational bots have come from the first version of Siri, the first version of Alexa, the first version of Google assistant and where they are today, light years. it is like, it used to be awkward to talk to Alexa.
If you talk to like these things now, it is like talking to a human. It is just unreal. So it is it's it's awkward, but you can get over that hump really quickly without even thinking about it.
Ben Ard (13:34)
Yeah, a hundred percent. Okay. These episodes are short, so we're running out of time. One last question. Let's jump into tactical advice. So in your role, working with your team, working with others, how are you capitalizing on GPTs and really getting the biggest bang for your buck across the entire org to really kind of capitalize on some of these GPTs and
Fred Faulkner (13:42)
Yeah.
Ben Ard (13:59)
Really make sure that you're using them the right way. Like, is there any tactical advice about, hey, if you're a team, here are some ways that you should use chat GPT, you know, custom GPTs today.
Fred Faulkner (14:08)
Yes.
Yeah, I got a couple of fundamental rules. First, human in the loop. Always. You just know. Whatever you're doing, human in the loop. Meaning, whatever you contact, create, get ready to edit, publish, human in the loop. Always. First and foremost. Second, come up with your use cases. So I think everyone gets intimidated by AI, specifically using tools like custom GPTs or whatever the plethora of platforms that are out there.
My biggest advice is come like sit it down as a team and just come up with use cases. What are the things that take up the most of your time? Data entry, data normalization, marketing content, drafting, brainstorming, whatever it might be. Pick one or two and then start to experiment with how they can help you improve that process further, faster, more efficiently. Then add the next one. Then add the next one. Do the before and after, right? Like before it took us this long. After it took us this long. Oh, that's a great improvement. Cool. Start adopting that and then iterate.
As you go along, right? And as you onboard new team members, make sure they understand these tools are here and here's a process we use to use them. That's like the easiest way to kind of approach it, just like snackable bites, then trying to solve like a huge workflow problem. Then there's some great tools out there. Like you can use a custom GPT or you can go use like a writer or a Jasper if you want to think about content. They have a lot of workflows and stuff kind of pre-built. The hardest thing for companies when it comes to marketers, if you don't have a good understanding of your
company's tone and voice, because I have a different tone of voice than you have a different tone of voice and how you then meld that together, the company's tone and voice by far, probably the biggest challenge is like, how do you consistently keep tone and voice in a company setting when it's not an individual is like probably the hardest, the hardest nut to crack, not impossible, but it is like, you have to really kind of get your other processes in place to then take advantage of AI. And I think that's the other part. AI will only take you so far if you don't have some of your already processes figured out. And if that isn't
clear, like data, data doesn't work in a garbage in garbage out. Same thing for AI. If you don't have your stuff figured out ahead of time, AI will make it better, but it will not make it as best as it can be because it's still going to fumble on certain types of things. So first of all, human loop. Always remember that snackable bites. Find your use cases and pick out your snackable bite versions first and then third. Get your processes and stuff figured out before you really start experimenting. Otherwise you're going to be garbage and garbage out. Those are like my three practical pieces of advice. And then
Try different platforms. You'll find ChatGBT is good at some things, Claude's better at others. You have to decide how much you're going to invest in different tools. But just go with your standard operating procedures and how you're going to then incorporate it into your workflows. That's the best way to take advantage of AI. you have to, because if you're not, your competitors are likely doing it and they're beating you to market on your ideas.
Ben Ard (16:34)
I love it. Fred, this has been a great conversation. I love the advice. I love all of that. It's actually going to change how I use my GPTs and all that kind of good stuff. I'll have to do more voice chats. I love that. But Fred, they're amazing. Yeah, this is great. I need to do it. I I do it so few at the time. I need to do it more often. I think this is great. Fred, for anyone who is listening and wants to reach out and connect with you online, how and where can they find you?
Fred Faulkner (16:46)
Boys chats, yeah. Way to go.
Yeah, thanks Ben.
Yeah, so you can find me at mcfadden.com, which is where I'm the SVP of marketing. But if you want to find me on LinkedIn, my handle is According to Fred, which is also my website, according to fred.com. So you can find me on most social channels that way. Happy to talk to you about AI and other kind of workflows when it comes to that at any time. So thanks a lot.
Ben Ard (17:16)
Love it. Perfect.
Fred, thank you again. This has been amazing. Really do appreciate it.
Fred Faulkner (17:20)
Yeah, sounds good. Thanks a lot, Ben. Appreciate being on the show.
