Mastering AI in Marketing: From Junior Colleague to Revenue Generator

Brandi Starr [00:00:49]:
Stephen runs the Sydney based marketing agency Taillist. Tailless specializes in direct response copywriting and Steven's team has developed techniques to use AI to research topics and audiences and then write copy that can be as close as 80% ready for publication. Steven, welcome to Revenue Rehab. Your session begins now.

Steven J. Lewis [00:01:16]:
Thanks, Brady. It's absolute pleasure to be here.

Brandi Starr [00:01:19]:
Yes, I am excited to have you excited for our discussion today. But before we dive in, I like to break the ice with a little woo saw moments that I call buzzword banishment. So tell me, what industry buzzword would you like to get rid of forever?

Steven J. Lewis [00:01:38]:
Passion.

Brandi Starr [00:01:40]:
Passion. That's an interesting one. So tell me, why would you like to banish passion?

Steven J. Lewis [00:01:48]:
Because passion is about you. It's not about your customers or your clients. Your clients don't care if you're passionate. If I go to the dentist, I do not care if my dentist is passionate. I care if my dentist is good at his job.

Brandi Starr [00:02:04]:
That is, that is a good point. Yeah. And it's one that I think I do hear the term passion overused. Don't hear it as often in business. So that is interesting because I do think that, I mean, I definitely agree with you that it is all about the quality of the output and not whether you're passionate or not or not. So I can definitely say for this conversation, we can put passion in the box and throw away key. Won't talk about it anymore.

Steven J. Lewis [00:02:39]:
That makes me very happy.

Brandi Starr [00:02:42]:
Awesome. So now that we've gotten that off our chest, tell me what brings you to Revenue Rehab.

Steven J. Lewis [00:02:50]:
It's what I'm seeing a lot of and helping my clients with, which is they've gone out and they've invested in these AI tools which present themselves as cheap. Right. You know, $30 a head per month. But if you've got a lot of heads, that's a lot of $30 if your people are not going to be using the tools. And what I see is people aren't helping people to use the tools. They're like, hey, you've got a tool. I've gone out, I've Got you this tool and they're assuming that people will use it and they will use it to its full advantage. And they're not.

Brandi Starr [00:03:28]:
Okay. And I know, I definitely can see that, especially with AI being integrated into almost every technology. And so as we go a bit deeper into that, I believe in setting intentions. It gives us focus, it gives us purpose, and most important, it gives our audience an understanding of what to expect from our discussion. And so what's your best hope for our talk? What would you like people to take.

Steven J. Lewis [00:03:53]:
Away on the absolute simplest nugget level? If you find yourself starting a conversation with your AI with one sentence, you're doing it wrong. That, that's, that's it. I would like people to check themselves and to know that if they see their employees or team members or whoever it might be typing into a chat, please summarize this document for me. And that is the most they think they can get from their AI. That's what I'd like people to realize. Hang on a minute, we're not doing this right.

Brandi Starr [00:04:27]:
Yeah, and I would definitely agree. And I think this is an important conversation. There's a lot of people who are, you know, fully invested in AI and doing research and coming up with some really sophisticated ways that you can do things. But as I talk to more and more organizations, there are still a lot of people who are really just trying to understand how to get started and really just, you know, I've had conversations, I mean, I have. They have a 23 year old, almost 24, and we were having a conversation and he was like, I know you love using chat GPT. He's like, I've tried to use it and what I'm getting out isn't beneficial. And I asked him, I said, well, what are you putting in? And it very much was that same kind of example, like here, tell me this or here, do this. And I was like, that's the problem.

Brandi Starr [00:05:24]:
You're not giving it enough context for what comes out to actually be meaningful. And so how do you, you know, what do you say to someone who is like, well, that's all they know. They're used to, you know, using Google or whatever. And you know, you kind of tell it, tell me this and it gives you something back. How do you articulate for someone what a good prompt looks like when they're really just getting started?

Steven J. Lewis [00:05:53]:
You've got to think of your AI as a colleague. So your AI, it's like a very clever but quite junior colleague. So if you had a person come in and sit in your office, and you thought, you know what? I want to get, you know, Gary to do this for me. What would you tell Gary? Because it wouldn't be, Gary, go away and summarize these documents for me. Because how's Gary going to know what you want from the documents? What are the documents about? Why do you care about the documents? What, ideally, would you like Gary to pull out to summarize from the document? So you would give Gary a lot of information. Hey, Gary, I've got four reports here. They're actually from our competitors. What I'm looking at is what are they planning to do in the next six months that we maybe should be planning to do or planning to be at least aware that they're doing in the next six months? Then Gary's got some framing.

Steven J. Lewis [00:06:49]:
Gary knows what the documents are, and Gary knows what success will look like from your point of view. So that's how you would start. How would I brief a colleague on this job who's not been in the game long enough, not been in my company long enough to infer that information?

Brandi Starr [00:07:09]:
Yeah. And I know I always say, think of it like a colleague who both knows everything and knows nothing. So it's like you have to give them all of the context as if they know nothing. But then the output and what they know on the back end is as if they know everything. And I do agree, like, having, you know, having a conversation with the AI is going to yield a much better output. And a lot of people feel like it's really awkward getting started. Started, or that when they're talking to a colleague, they don't have to go that deep. They don't have to give, you know, I don't have to tell Gary, give me a word document with, you know, bullets and highlighted headings.

Brandi Starr [00:07:58]:
I can just say, here's the outcome. And so what, you know, what input do you give to people who are kind of struggling with that middle ground of, you know, this is more than what I would have to say to a colleague. How do I know what to say? Um, you know, I'd love to hear more of your thoughts and advice in what you give to clients.

Steven J. Lewis [00:08:19]:
I think if somebody came to me and said, I. I wouldn't have to tell my colleague that, I'd say, well, if this was day one for your colleague, I think you would. Because even if your colleague was in the habit of summarizing documents for their previous employer. Now bear in mind my. My brief is. It's a young colleague, it's an inexperienced colleague. So it's not someone who's had a load of job experiences doing the same job for seven other companies. It's, they probably haven't had a job doing what they do now for a company like yours.

Steven J. Lewis [00:08:50]:
So you probably would have to say, in our company, the style is this, this is what we do. We use Word documents. So to give you an example from my own career, mostly I've worked for myself. So I've had some experience working in large companies. And the last time I worked in a large company, I never got over how attached they were to PowerPoint because I'm a writer and, and I understand that when Microsoft created products, they called Microsoft Word Word for a reason, because it's the one you use when you have words. PowerPoint is the one you use when you have pictures and a few words. But this company that I worked for, everything had to be in what they called a deck. I'd never heard the term deck before.

Steven J. Lewis [00:09:38]:
I know deck is a very common term and now we have to say deck all the time. But my point is it's not logical to put things in a PowerPoint when the things are words. The software, it's not logical because the software doesn't allow easy editing, easy fixing, easy flow, easy spell check, any of those things. PowerPoint is terrible at all the things that you want it to do most of the time. When certainly this employer of mine wanted a PowerPoint, so it wasn't logical. So as a new person coming into the business, and I was well over 30, well experienced, that was completely new, people would have to say to me, no, Stephen, even though this thing is only words, we need it in a PowerPoint. So that colleague of yours, I think people are kidding themselves if they think they wouldn't have to give a young colleague a lot of I want bullet points, I want highlights, and I want it in a Word document. I think they're dreaming and I think their expectations, both of their young human colleagues and of their AI is probably unrealistic.

Steven J. Lewis [00:10:42]:
And that means if you, if you're the CRO, who's, you know, looking at where, where are we getting our bang for our buck or our cmo who's like, where? Why am I getting, you know, no efficiencies from this AI, it could well come down to people giving bad briefs to, to the AI, and that's going to be expensive ultimately because you could get much better results from better briefing of your AI and, or you could just stop using the AI if you're not getting results and save Yourself money that way.

Brandi Starr [00:11:12]:
Yeah. So I want to shift gears a little bit and let's talk about workflow because you talk about, you know, having techniques to leverage AI to go from researching topics to writing the copy to, you know, getting it close to about 80% ready for publication. And I think another mistake I see a lot of people make as they are getting started with AI is they think everything happens in one step. Like you just say, hey, here, write a blog post on this thing and magically it's supposed to come out with something amazing. And so I'd love to hear your take on, you know, what are the steps? Like what's the workflow of figuring out? And we'll just use developing content as a great example because that's something that most marketers are trying to tap into AI for. What does that workflow look like from start to finish?

Steven J. Lewis [00:12:07]:
What you have to remember with the, with the chat based AIs like ChatGPT is that they have something called the context window, which you know all about. Brandy, which is, you know, every time ChatGPT answers question, a question, let's say it's the tenth thing that you've said in the chat before ChatGPT gives you that tenth output, it goes back and reads your first question, it's first answer, it's your second question, and so on all the way back. So you and I could be having a conversation about, you know, what the weather's like in Georgia right now versus in Sydney where I am, because we're having winter and you know, you're not. So you, when we move on to chatting about, ChatGPT will forget all about that, you know, lovely conversation about the weather, chat GPT won't. If you start a chat with chat GPT talking about the weather and then you want to go on and talk about your marketing, ChatGPT will be using all that weather information at the top of the chat to inform the marketing stuff. So in terms of a workflow, you've got to think in terms of what I call chat hygiene. So there's probably a research component to putting your content together. That research component might involve many steps.

Steven J. Lewis [00:13:19]:
I want to have a look at this blog post. How good is that? What do they talk about? Okay, now I want to think about some topics. Now I want to do whatever those should probably be happening in different chats. So if, if one's mindset is I can write a blog post from research, from first prompt about research to final draft of the blog post in one chat that's going to give you a really bad result. So probably if you're producing something important like a piece of content that's bigger than a LinkedIn post, if you haven't had multiple separate chat engagements with your AI, again, you're probably making some basic mistakes about how the AI works. So you'd have all of your research happening. You're probably in several chats where you, your job is you're pulling out the good stuff and you're collecting that somewhere else. Then when it comes to drafting, you're probably pulling out that good stuff, putting it into Chat GPT.

Steven J. Lewis [00:14:19]:
Maybe you've got a couple of great paragraphs, you're working with chat GPT on those paragraphs, then you're starting another chat to talk about different paragraphs. And then finally, when you've got something you like put together, you should be loading that into Chat GPT or your AI and saying, what do you make of this? Where could I make it better? I've got this draft. This is my audience. What could make it better? So it's a really, as you say, it's a workflow, it's very structured, but it's happening over multiple chats.

Brandi Starr [00:14:54]:
Yeah, And I definitely agree there. One thing that I have done that I have found success with is when I've gotten to the end of one component. So let's say I, you know, I'm working in one conversation to do some research, and as I'm going along, if there's a piece that I want it to hold on to, I will say I really like this and want to keep it for the output. Make sure to hold on this for the final. And then once we get to the end, I will actually tell it, hey, I'm going to engage another conversation that I want to pull out all the nuggets from this one. Can you summarize all the things I liked and any key notes or highlights from, from my input that would be good for the next conversation? And then that usually will produce not only a summary of all the things I like, but where I have found real value is it will often really summarize my thinking and direction and kind of that back and forth, brainstorming in a way that I've not been able to articulate. So when I've worked, you know, I've worked with writers and things in the past, and I know I have always struggled going between steps in being able to synthesize what I'm really feeling and what the input is that they should carry forward. And so I'm seeing That AI will spit that back to me.

Brandi Starr [00:16:24]:
I can copy that last response to then go. Okay, now we're going to start drafting this. Here's all the things. Give me an outline that you think would be great and, you know, go back and forth with that. And that's another technique that I have seen. Are there any of those sort of, like, key, you know, this is something I've found that has really enhanced this process that you can share.

Steven J. Lewis [00:16:51]:
I think using AI as a sounding board, as you are doing, there is something people, people miss. I think AI or certainly in terms of ChatGPT, it's been sold as a generator of content and it's misunderstood by a lot of people because it generates, left to its own devices with inadequate prompts and bad briefs, it writes terrible copy. And that it's really easy then for people to dismiss it. Oh, Chat GPT is bland. It writes rubbish. Why would anyone use it? Which is like the people who want to tell you that social media is just people taking photographs of their lunch. And, you know, the Internet is just full of people who, you know, post pictures of their cats. And you, you're like, yeah, okay, like, if you just want to dismiss the entire Internet and all of social media, very all right, your view is boring.

Steven J. Lewis [00:17:45]:
But. But, okay, fine. It's just dinner photos and cats and, and chatgpt is boring. It's just frankly, lazy thinking, right? It's people who haven't investigated properly how to make ChatGPT do something good. As I say to people, it's like if, if I broke into a Formula one garage and jumped in the cockpit of a Formula one car, I wouldn't even know how to start it right, let alone get it out of the garage and whizzing around the track. I don't know how to do it. Does that mean a Formula one car is rubbish? Or does that mean I'm not a good driver of a Formula one car car, right? And Chat GBT is a Formula one car that not only can write if you prompt it correctly, but can help you with your thinking. So a lot of the prompts that I write for people, like in my course, make Chat GBT your cmo, right? That is a course that is A lot of it is about getting chat GPT to help you with your thinking.

Steven J. Lewis [00:18:43]:
It. It's 30 days, an email a day with a prompt a day where you're not writing anything. Like, nothing that you learn in that course is to write something that your ideal client will ever see. It's all about getting Chat GPT to consult to you. And that, I think, is something, to answer your question about something that I think is brilliant that people miss is ask. Ask a properly briefed ChatGPT. So by that I mean, if you've trained it to be your CMO and you say to it, I'm planning to write this, or I'm planning to say this at the next meeting, what do you think? It can give you brilliant insights and show you things about your own work that you have not seen for yourself, which can be quite confronting.

Brandi Starr [00:19:28]:
Yes, and I definitely have had some of those uncomfortable moments where I've asked Chat GPT to be brutally honest about either something I've written or my perspective on something. And, you know, there's. I've had a couple, like, oh, clutch the pearls moments where it's like, I thought I was on the right track and, you know, it was kind of like, everybody's saying that, or, you know, it's like, this is overdone. Like, here's, you know, five articles talking about the same thing. And I'm like, oh, okay, well, let me figure out how, you know, what is. How is my perspective different? Or, you know, even being in consulting, in some cases, I get pulled. Pulled in to help on, you know, developing proposals or, or scope for clients. And in some cases I'll ask, here's who my audience is.

Brandi Starr [00:20:20]:
How do you feel like this is going to be received? And in some cases, you know, it'll say, well, I'd be confused by this, or this does this component doesn't look worth the cost, or, you know, this doesn't seem like it would address the problem. Like, and other places, it's like, no, this would be great. I definitely want to, you know, have further conversations. And so that is like. And I think that's the thing I've been leaning into most with my team is getting Chat GPT to be that thought partner with you, to be able to, you know, pull in research or what others are talking about to ask questions. That's another thing I'll always do is say, like, what do you need to know from me in order to give me a good opinion? Or what should I be thinking about in order to X, Y and Z. And it quite often gives me things that as long as I've been in my career, as savvy as I am in what I do, it often pushes me to think about angles and topics that just had never even crossed my mind. And so I definitely think that's a place, you know, where a lot of our listeners are, you know, most of our listeners are marketing leaders, VPs, CMOs.

Brandi Starr [00:21:41]:
And so I'd love to hear you dive in a little more around, you know, not without, without giving away your secret sauce, but help, you know, those listening to understand more. If they've never leaned into using AI to help them think through things, how can they go about really just thinking about that as a use case differently?

Steven J. Lewis [00:22:05]:
Well, in my course, which is about making chat GPT your CMO, and to all the CMOs listening, nobody is pretending that this is actually going to make Chat GPT your actual cmo, but it's a consulting partner in the realm of marketing. And what your CMO is, the one that we build in the course is, it's what's called a GPT. And a GPT. So you've got ChatGPT, which has its training data. So ChatGPT, as you were saying earlier, knows a lot of things, right? Like, it might not be an experienced colleague, but my goodness, it has read an awful lot of material, right? It knows a lot of things. When you build a GPT, what you're able to do is layer your own information over chat GPT's existing information and you're able to give that GPT special skills. So if I, if I can break that down, when I say give your, your Chat GBD special skills, what I mean is, theoretically, if you said to Chat GPT, please write an invitation to a webinar, Chat GPT should think for itself, right? Okay, what skills am I going to need? I'm going to need some marketing skills. I'm going to need some copywriting skills.

Steven J. Lewis [00:23:16]:
I'm going to need to understand the topic of this webinar and the audience for this webinar. Let me get that information together and then I'll apply it. ChatGPT doesn't. If you go to ChatGPT and say, Write me a webinar invitation, that's literally the only thing you say to it. It'll just start writing. It won't say, hey, Brandy, what's the webinar about? When's it happening? Who's it for? Who do you want to come? Like, it won't ask anything, it'll just do. So the instructions that you give a GPT, which is the, this layer that runs over ChatGPT, would be okay. I want you to have these marketing skills, I want you to have these copywriting skills.

Steven J. Lewis [00:23:51]:
I want you to have these abilities. And so in, in the course, that's about a thousand words of instructions that I Give people to give their cmo. Then you can add information like, who are my competitors, what do I know about them, what do I know about my ideal clients, et cetera, et cetera, et cetera. So that when you go into a chat with ChatGPT and you say to it, I'm going for a meeting with a client, what should I be asking? Chat GPT has all of this very specific information about you that isn't in its training data, that isn't part of its general knowledge. So then its answers are being filtered through an enormous amount of structured information that you have given it. So people who take the course, for instance, say to me, I'm getting information about my ideal client that I've never been able to get before. Like, I've interviewed my ideal clients, I've asked them to take surveys. I have never got this depth of understanding.

Steven J. Lewis [00:24:49]:
It's that kind of information that you pull out and then you have Chat GBT answer through. And that gives you a richness of response from Chat GBT that you're just never going to get by saying, I'm writing an email to mothers aged over 35 who live in, you know, the continental U.S. you're just never going to get an answer from Chat GPT that gives you the same richness as if you've structured the data first. And then you'll find you almost won't want to do anything without first asking Chat GPT what it thinks of it. So you've just got that person sense checking you all the time. You don't have to agree with it, but you can think, okay, I've since checked that and I'm okay with my decision, even when you don't agree, which is exactly what you do with a highly qualified human colleague. Thanks, Brandy. I appreciate your advice.

Steven J. Lewis [00:25:44]:
One of us has got to make a decision. I'm going to go with your view, or I'm going to go with my view, whatever it might be. But you've had the good sense to check it with an expert first.

Brandi Starr [00:25:53]:
Yes. And I, you know, the, the custom GPTs, when I finally learned how to leverage those, has been the biggest game changer. And, you know, we're a marketing team of three, but with all my custom GPTs, we're a marketing team of nine with three more air quotes being hired, which are three more GPTs that I've got to build. And each of those are trained on a really specific thing. So, you know, we've even gotten where we have one that checks our tone. So we have trained it on how to sound like us. And before anything goes out, whether it's just social content, a white paper, whatever it's. Does this have the Tegrada tone talking about the icp? I built one that is our icp and it is, you know, we check it and say would you respond positively to this? If you were comparing us to a competitor, would this help move you in the right direction? And it gives us brutal, honest feedback of like, eh, I wouldn't really care about that.

Brandi Starr [00:27:05]:
Like that sounds great, but that's not important to me. And I have noticed just in some of the emails that we have leveraged that ICP GPT to check against our response rates have increased quite significantly because the language is speaking the language of our audience. And so now it is to the point where certain times I'm like, I'm not doing that until I check, you know, we've named him Carl. I was like I'm not doing that till I've checked with Carl. And I refer to Carl so much that people on our team, you think Carl was like a real person that we were talking about because they'll send me something. And I'm like, well did you check with Carl? What did he say? And you know that that is like that has been the game changer.

Steven J. Lewis [00:27:58]:
Well, it's so interesting you say that because one of the things that I found in the course that people are emailing back to me is I haven't suggested all of my GPTs have a name, but I haven't suggested to anyone taking the course that they name their cmo. They all have, whether it's a name like Carl or. I had an email yesterday from a lady in Spain who's named her GPT Maverick, you know, so you can't. As I, as I said to people when I realized they were named them, I wrote down, I said you can't have that much help on your business from somebody without wanting to give them a name. It just seems wrong to call it. It's when it's been so helpful.

Brandi Starr [00:28:45]:
Yes. And for me the, the reason I started giving them names was the first one that I built was before I learned not to try to make a GPT do too many things at once. And I gave it a generic name. And then over time I started to see the flaws in that GPT and I was like, oh, I've got to build multiple. But then it was like, how do you distinguish between, you know, all of the GPTs? And I'm like, well, when I'm talking about People on my team, if I say, oh, you know, Sarah's responsible for that, everybody knows who Sarah is. They know what Sarah's job is. So they know what to go to Sarah for, when and where in the workflow. And so by giving them all names, we're able to have that same conversation the same way I can say, oh, that's something Sarah should help with.

Brandi Starr [00:29:39]:
I can say, oh, that. You know, Carl should help with that. Annie should help with that. And yeah, it is funny, because we do have conversations about them as if they are real people on the team and they're a part of our workflow, the same way, you know, our actual colleagues are.

Steven J. Lewis [00:29:58]:
And it will help you to remember to talk to them like they're a colleague. Because when I'm talking to Gene the copywriter, who's my. My copywriting GPT, I'm remembering Gene as a person. Gene needs information. I can't sort of make assumptions that Gene knows everything that I know because I. I haven't told him or April Avatar Hunter, who's, you know, very good at coming up with those ICP Personas because that's what she's been trying to do. She's got a name, and they've all got kind of fun names for fun reasons, and, you know, it. They are enjoyable.

Steven J. Lewis [00:30:33]:
This is one of the things that is so frustrating, right, for you. I think I'm imagining. I'm guessing you can tell me, and for me is you can't really convey to people until they've tried it how enjoyable it is to work with a properly trained GPT. It's not like typing something into word. It's fun.

Brandi Starr [00:30:53]:
It is. And, you know, we are a fully remote team. And so I don't. Other than my son, I don't always interact in person with anyone. You know, I'm on meetings, and so there are some times where I have brainstorming sessions, and I'll even be, like, drawing on my whiteboard and then, you know, like, typing to my GPTs. And I'm like, okay, Carl, what do you think about this? Okay, Rachel, like, how would we then. Because we have repurposed. Rachel, how would we then take that and repurpose it? And it's like, okay, you know, now let's go back to we got SEO Sam.

Brandi Starr [00:31:30]:
Like, okay, Sam, how. Where do you weigh in on this? And so, like, I feel like I'm in a meeting with all these people. And for me, often it's like, super late at night, and I'm like, we're vibing. And I'm really in a room by myself. And it is a lot.

Steven J. Lewis [00:31:48]:
Somebody listening to this brandy is going to think you sound nuts. But if you're listening to this and you think brandy sounds nuts, try it. And then you will, you will not think it's nuts. You will be converted.

Brandi Starr [00:32:03]:
Yes, I agree completely. And so thinking about it, talking about our challenges is just the first step. And nothing changes if nothing changes. And so in traditional therapy, the therapist gives the client some homework, but here at revenue rehab, we like to flip that on its head and ask you to give us some homework. So for those listening who recognize that they haven't even dipped their baby toe into the AI world, where do you recommend that they start? What's that one thing?

Steven J. Lewis [00:32:37]:
I've got a five minute course on my website. It's free. You have to give me your email address. So, okay, it's not free. I will charge you one email address in order to take it. Right. But it's a video, so you can watch part of the video first. It will show you how to get chat GPT to write in your tone of voice exactly like you.

Steven J. Lewis [00:33:00]:
Because I think that's one of the obstacles that people have to chat GPT. It's the cat picture, it's the dinner p. Oh, chat GPT is so bland, I can't use it. It sounds rubbish. Well, what if you could make it sound exactly like you? So I've got a prompt that will divine your tone of voice and give that back to you so that you can use that as a prompt to ChatGPT. So you could say to ChatGPT, I want to write an email in this tone of voice. And then, and then it will do that. What it will also do as a sort of side benefit of a benefit already is it gives you a bit of a review because what it does is it will read, say, your website brandy, and it will give you the tone of voice that it's determined that website is written in.

Steven J. Lewis [00:33:44]:
But you can also take that as a review because if you get that back from an objective third party and it says, I don't know, this website is very formal and you think might, I don't want to be formal. That's not how I want to come across. For example, you can treat it as a mini review, like a theater review. So for instance, I trained it on my email newsletters and it was very flattering. It's like, you know, Stephen's quite funny, but he doesn't let that get in the way of his knowledgeable approach, but, you know, he uses metaphors to make his part. I'm like, yes, that is me. That is exactly how I want to be. Thank you, Chat GPT.

Steven J. Lewis [00:34:22]:
But equally yesterday it was kicking my backside every which way to Sunday by reading it and saying, actually, this is, you know, too formal. This is, you know, this is reading like a, a statistical fact sheet, not a very human piece. And I'm like, do you know what? I have to be honest, because I was writing something for a doctor, so I got deeply immersed in the research and the science and I'd read, you know, several medical papers and I realized I was letting all of that infuse the copy. So instead of writing in a human way to this person's patients, I was writing because I was very pleased with all the knowledge that I'd acquired and I wanted to share it. So, yeah, it kicked my butt. So it's a really good, it's a really good prompt. It's free if you go to Taylist, which is T A L E I S T as in telling tales, Taylist agency forward slash mini M I N I. Because it's a mini course.

Steven J. Lewis [00:35:20]:
It's a five minute course and it will teach you how to get ChatGPT to write exactly like you.

Brandi Starr [00:35:26]:
Yeah, and I definitely can say that that is a great starting point is to figure out how to get it to write in your voice because that is definitely the biggest game changer. When I worked on my voice paragraph, I fed it some of my LinkedIn posts, some of my emails, and then I had it write some social post. And I also wrote some social posts and sent them to my team to see if they could take tell which ones were me. And they couldn't identify the difference. Like they all sounded the same. And I was like, okay, that means I nailed it. If, you know, the people who read everything that I write can't tell the difference, then, you know, I know that I've gotten that right. So for, yeah, I, I definitely agree for anyone getting started, that is a great place to start because it will then allow you to sound like yourself.

Brandi Starr [00:36:24]:
Because I will agree that tone is probably my biggest challenge with AI. Like, you know, we all have the orchestra metaphors, which as much as I love a good metaphor, I now hate orchestra metaphors because of Chat GPT. You can always, you know, you can always tell that that that's one of them. So we will make sure to link over to that. So check the show notes wherever you are listening or watching this podcast. So, Stephen, I have enjoyed our discussion, but that's our time for today. So before we go, tell our audience how they can connect with you. We've already given the shameless plug and had the website, so we're going to link to that.

Brandi Starr [00:37:08]:
But if there's anything else that you want to share with our audience, now's the opportunity.

Steven J. Lewis [00:37:14]:
I'm all over that website and if communications are your thing and and you want to know how to be doing it better, if you visit the website, you'll be asked a thousand times to sign up for my newsletter. So that is the best way to take advantage of some of the tips that I'm able to share.

Brandi Starr [00:37:31]:
Perfect. Well, again, we will make sure to link to that. And this has been such a fun discussion. I am all in on leveraging AI. I don't think there's a day that goes by that me, ChatGPT and Claude don't have spend some time together. So again, thank you so much for joining me.

Steven J. Lewis [00:37:51]:
Thanks, Brandy. It's been an absolute pleasure.

Brandi Starr [00:37:53]:
Awesome. And thanks everyone for joining me. I hope you have enjoyed my conversation with Stephen. I can't believe we're at the end. Bye. Bye. You've been listening to Revenue Rehab with your host, Brandi Stahl. Your session is now over, but the learning has just begun.