Ep #130: AI Tools for Fractional CMOs

Transcript:

00:00:00 Casey: In this episode, I’m going to talk to you about AI and what AI tools you should use to save time and deliver a better work product as a fractional CMO. Let’s get into it.

00:00:11 Casey: Marketers of the world, why do we work hard to solve small problems? Why do we reinvent ourselves and our clients over and over? And why are we giving away marketing strategy for free? With advancements in AI, we’re all seeing the marketing department shrink from the bottom up and companies need you to serve them as their fractional chief marketing officer. It’s time to solve bigger problems and bring home a bigger paycheck. It’s time to create the lifestyle we deserve and to make a greater impact. This is the Fractional CMO Show and I’m Casey Stanton. Join me as we explore this growing industry and learn to solve bigger problems as marketing leaders. The Fractional CMO Show is sponsored by CMOx, the number one company to teach you how to attract, convert, and serve high paying fractional CMO clients on your terms.

00:01:07 Casey: Hey welcome, it’s January 2026. Can’t believe we’re here. Excited to share with you some of my thoughts on A &I. And let me give you a quick arc. about 14-15 months ago, I finished my second book called Find Your CMO and it was kind of at that time, as I was wrapping up the book, where I was really kind of considering like, do I stay in fractional CMO work? Do I be a fractional CMO? Or do I go do something else? And as you’ve seen, like the proliferation of AI and all the stuff that’s happening, there’s an exciting time. It’s kind of like a couple years ago saying like, ooh, should I get into alternative crypto coins? Or before that, should I get into Bitcoin? Right? It’s like we’re in these different eras and I gave myself the time and the space to consider if that was a thing for me to do or not. 

00:01:56 Casey: As you can probably tell, I’m like a little technical. I’m technical enough that I can break something and then committed enough to spend all weekend trying to fix it. Right. That’s kind of where I fall on my technical skills. I’m not all that talented, but I do like to dive into technical stuff and I think I can figure some stuff out. So I really gave myself that opportunity to say, don’t do the fractional CMO thing, or do I go and kind of jump into this whole like wild world of AI and rags and LLMs and you know, all the stuff.

00:02:24 Casey: And what became very clear to me was that the role of the fractional CMO is the role. Like that’s the one for me, that’s where I want to be. And I love leading CMOx. I love helping all of these other great CMOs who are building their fractional CMO practice. I love being a business owner. I love having our sister company CTOx and watching that grow. But at end of the day, I want to be a CMO. That’s really where my… I guess like my commitment lies. That’s where my leverage is. That’s, that’s where I’m having the most fun too, because I get to show up and I get to make things happen for clients. And I maintain a book of a couple clients and I don’t really want any more. And my clients are all very long-term commitments, multiple year long commitments. 

00:03:09 Casey: And the whole goal is like, I want to build something big with and for the clients. I want to build a real tangible marketing department. And I want that to help the business scale. And I want to make a bunch of money doing it. And I want my client to as well. And I think back to this idea that I had years ago when I joined a marketing agency, we had these projects that we called core projects. And a core project was a project that the agency would take on, everyone would volunteer their time for, and only if it was successful, did we make money. So there was like the traditional day-to-day work where I was paid hourly. I was a contractor.

00:03:42 Casey: And, you know, I started at like 20 bucks an hour and I moved up to like 50 bucks an hour. And that’s just like the life I lived as, as a contractor there was contract work somewhere, 30 hours a week, 40, 50, 60, you know, kind of depending on what the week was like. Then there was a core project that would come in and it would be like the business owner would meet someone at a mastermind and then that person would want to do a project. And the business owner of the agency, he would negotiate a thing where it said, Hey, if we do this and make this come true, we’ll take 50% of profits, but we’ll, it’ll cost you nothing upfront. So all the risk is on us. And we would take on projects like that. 

00:04:15 Casey: And I took on every single core project that came in. That was like my commitment. I was gonna take on every ounce of risk I could because I was very risk tolerant. I was young, I didn’t have a family. I was living in New Orleans, pretty inexpensive city. I biked around everywhere. If I went out, I went out with a six pack of beer in my backpack. It’s not like I was living an expensive lifestyle. So I could afford to take that risk.

00:04:40 Casey: And what I found was that pretty much none of the opportunities paid out. Two of them paid out over time. And one paid out nicely, you know, maybe it paid me an extra two grand a month for maybe a year. And the other one, I think it paid me maybe five grand as a single time thing. Not big money, not money that I can retire on none of that. And I was taking on all of this risk. And like, I still want to take on risk. But I also want to make a lot of money while I take on risk and I want the risk only to pay me even more money. 

00:05:12 Casey: Does that make sense? Right? Like I want a baseline of making a great living. If all of the risk I take doesn’t turn out, I still want to make good money. So the idea here is that moving to the fractional CMO role is I get paid well, and then I can take risk on top of that and get paid even more. So that was the original commitment, right? And that’s what I was looking to do was build a book of business that paid me well that gave me the opportunity to pay me really, really well. That’s kind of what I set out to do. 

00:05:46 Casey: And I want to do it with as least risk on the monthly like every month, I want to hit a certain number of dollars in my pocket, no matter what just done like I want that cash coming in easy cash flow, I can budget for it, save some of it, you know, do all that stuff. And then all the upside, you know, that goes straight into savings. Okay, so with that idea in mind, I shifted back to the fractional CMO focus. And over the last year, I’ve taken on new clients, I’ve added four clients, and I wrapped one, we worked together for maybe nine months. And now I’m in long term relationships with these three clients, and we’re having a great time and uh things are looking really good for everybody. 

00:06:29 Casey: So there’s AI tools that I want to talk about that have set me up for success along the way. But there’s also some caveats here that I want you to know. So I think the first thing I want to say before we even get into the tool is that I want you to be a Renaissance man or a Renaissance woman. That’s what I want you to be. I want you to identify as a Renaissance man or a Renaissance woman. I want you to identify as someone who understands the art of the old, of how to do things manually. Like, do you know how to change the oil in your car? Like, do you fundamentally understand how to change a tire on your car? Do you know what setting the torque correctly on the lug nuts is, like, do you understand this stuff? 

00:07:13 Casey: If you do, it’s fine. You can have someone else do the work for you, but you know what they’re doing. There’s no magic. There’s no black box. To be a Renaissance man or woman is to understand how to do the thing how to write persuasive copy how to write good content. But you have to fundamentally know how to do that. Because if you lack the knowledge and how to do it, you’ll never be able to get AI to do it for you at a level that’s appropriate.

00:07:37 Casey: So I want you to commit to being a Renaissance man or woman. This looks like just doing the labor, the inefficient labor of doing stuff, of just like writing stuff, of trying stuff, of getting inside of the ads manager and poking around alongside your marketing technician and learning along with them. Like every single time I log into the ads manager, which isn’t all that often, there’s always some new AI feature from Meta and like they’re going to offer you headlines and they all kind of suck, right? Okay. But that was maybe six months ago. And now the headlines that get offered when I write an ad, they’re actually pretty good. Three of this, three of the five that they give me in the first round look good. And I actually, sure. Those are good enough for us to run. 

00:08:20 Casey: I want you to get in and play. So you have to know how to do it. And then you have to go play with the platform and see what AI is available there. And I think that you’ll find that maybe there’s a nice middle ground, but it is the people who are saying, I’ve never written persuasive copy before. I haven’t studied the art of persuasive copy. I don’t know the Gary’s, right? I don’t know the Gary Halbert, Gary Bensavanga. They don’t know these names. They don’t know the name Claude Hopkins. They don’t know the name Dan Kennedy. They don’t know John Carlton. They don’t know these like David Deutsch, these great direct response marketers. Paris Lomfropolis, he wrote the forward to my book, The Fractional CMO Method. Like these great, some of them living copywriters, they don’t understand the way that these people think how their copy is different, how Paris writes versus how Halbert writes versus how David Deutsch writes. Like they’re very different styles of writing. 

00:09:15 Casey: And I want you to have that sense of understanding and taste. And if you lack it, AI won’t fix it. AI will give you a bandaid, it’ll feel fine, but you’ll be developing a lower quality work product. So I get some heat around this all the time. When I say that, I think that the kind of core understanding of marketing kind of comes from the lens of the copywriter. I think copywriters, people who write persuasive copy are by and large stronger marketers than people that come in with a brand focus or people that come in with, you know, more of a MarTech approach. The copywriter understands the prospect, what their needs are, what their desires are. Remember, no one desires someone’s product or service.

00:10:01 Casey: They desire something larger than that. They desire freedom or power or beauty or youth and only the product or service can help them get that. That’s the line that you have to draw. Like these fundamental understandings of copywriting and how copywriting works are so critical to understanding how to be a chief marketing officer and therefore a good fractional chief marketing officer. So you have to understand that fundamentally. So the first thing that I think you should do is before you invest in an AI tool, I think you should invest in a book.

00:10:32 Casey: Like when was the last time you bought a good marketing book? I’m not talking about just like the pop books of today. You know, like Chaldini’s Influence, great book. New work kind of poppy, but still great. Influence is awesome. There’s a collection of books like that. But just find like Gary Halbert’s letters, the Gary Halbert letter, go to that website and read all of his past letters. Read one of them a day. Read the work of John Carlton. Read the work of David Deutsch. Read the work of Paris Lempropolis. Read the work of these people who have just done a tremendous job of writing persuasive copy and start to understand it and internalize it and maintain a swipe file. I’m working on a project and we’re advertising a travel offer for simple terms. And what does the website look like? Like it could look like anything we want, right? We can make the travel website like a boring page. We can make it look like a typical travel site. What we decided was we wanted to go after music events like big festivals.

00:11:30 Casey: And we looked at the best festival websites and went on the internet archive, archive.org and looked those up and pulled together a swipe file of fun, high energy, exciting landing pages. And then wrote copy on that and then used AI to like leverage what we’ve written and write something maybe better. Okay. But the steps along the way are like understanding marketing is critical as the first step. And that’s you being a Renaissance man or a Renaissance woman. You have to understand fundamentally how ads work.

00:11:57 Casey: And they’re just small nuances. Like someone recently proposed to me a postcard to review. Say like, hey, Casey, can you review this postcard and if you like it, then we’ll get it sent out. And I looked at it and like it was pretty and the headline was good and the imagery was good. And it just didn’t sit right with me. And it took me a minute, right? And I came back to it a little bit later that day. And I was like, I see what’s missing. A distinct call to action with timeliness. Like that’s what it was.

00:12:24 Casey: By all measure, it was an adequate postcard for most people, but it was missing the mechanism, like a little coupon cut that says expires on this date so that people actually take action and do something. I just needed the space to kind of consider that. If I asked ChatGPT, he would have said, Hey, that looks like a beautiful postcard. Congratulations. Right. It’s going to be, how do you say it? Single-fainted? Like it’s going to just like tell me that I did a good job. I don’t want that stuff. I want to think critically. So you have to fall back to this base understanding of marketing. And you should be a student of marketing.

00:12:54 Casey: You should be a student of direct mail. You should be a student of uh billboards. You should be a student of ads on Facebook. You see a good ad on Facebook, save it, copy it and save it. Where do you save it? Well, probably somewhere better than just like Apple photos. You should probably then move that into a Google drive or maybe use Obsidian or something like that. And you can organize things. Maybe you’re a Notion person, whatever. Save that stuff. You want to be able to generate a swipe file that you can reference later.

00:13:24 Casey: So before we talk about AI tools, what do you provide that the AI can’t provide? You provide taste. That’s a huge thing. Taste is massive. Your taste of like what looks right, what feels right. Like my wife, she’s really into interior design in the winter and in the summer, she’s really into flowers. And right now she’s on uh a kick with interior design and she’s got some great ideas.

00:13:51 Casey: I can run AI and say, Hey, what color is like a good color for this entryway? And we have this type of sun, you know, we in our front entryway, we have like a center hall colonial. So our house faces oh south. And, you know, with that, like, what’s good for sun sunlight, how does the color cast it? Is it like, is it like a blue color? And therefore, you know, what good paint color should we use? I can come up with those things and ChatGPT can give me an answer. My wife has taste, like, her taste is significant. I lack taste. In interior design, I just lack it. I just don’t even know what’s right. I can just follow something along and it’s all built together. And I’m like, yeah, that works.

00:14:34 Casey: You know, I don’t have a sense of taste. When it comes to marketing, I have a significant sense of taste. I think I’m very good at identifying what tastes good, like what looks good. I’m a stickler for things like closing an email. You say a salutation at the bottom, like a sign off, like sincerely or, you know, talk soon or, you know, to your success, comma. And then I like to do three enters before I put what I’m signing off as my name or the client’s name or whatever. And I like that for spacing because it harkens back to a handwritten letter where there’s enough space below the letter to sign your name. And then maybe you print it below that. And that feeling just works for me.

00:15:18 Casey: And if you lack that it feels like someone just like sentence enter sentence enter sentence enter sentence enter send. It just it just like lacks that feeling. There’s feeling that you might have like a sense of taste for branding. Right? I can go to ChatGPT and use Dall-E or one of these other stable diffusion generators and come up with a cool logo or image or whatever. But it lacks taste right? Like, if you are going to work with imagery, become a student of design, become a consumer like a student, a Renaissance man or a woman who understands composition. It’s like a fundamental in photography is composition. There’s like the rule of thirds. If you don’t know the rule of thirds, just learn them. It’s simple, right? 

00:16:03 Casey: Like learn these little things so that you can prompt more effectively. So taste is one of the things that we offer that AI doesn’t offer. The second one is discernment. Like we know what to share and what not to share. We know how to share it. Oftentimes AI can be like overly aggressive or under aggressive. It’s never in that Goldilocks point of actually meeting the prospect where I want to meet them. And there is such a difference. If we take the words, you look great in that dress, and we just emphasize any word in there we want, changes the meaning. You look great in that dress. You look great in that dress. You look great in that dress. You look great in that dress.

00:16:47 Casey: Like it changes the meaning. That’s discernment. How do you look at something and discern what works or what doesn’t? That takes a human element today. So if you lack that discernment, you need more time under pressure, like time under weight. You need to just be like getting your reps in with your clients. That means you might be writing some copy or reading some books or just like learning to be a better marketer. Discernment matters. Because I can go to Claude and ask them to write a webpage for me.

00:17:16 Casey: And sometimes I’m just like, good enough. I’m just my discernment is like, I’m just gonna like not even care. It covers enough of the basis, just send it to print. It’s very different than doing it right and caring and making sure that the work product is high quality. So you have taste discernment. And the third thing is experience. Robots don’t have experience. You know, they have a large language model. They have a statistical significance, but they don’t have experience.

00:17:44 Casey: The stories of my experience are potentially interesting. When I talk about my wife being interested in interior design in the cold months and growing dahlias in our backyard in the warm months, like you see that we have some experience there. She has some experience there. You’re also relating to that story. You’re like, oh, I don’t do the interior design, but I do do something else. I don’t do dahlias, but I do enjoy, I don’t know, grass, whatever.

00:18:12 Casey: Experience matters here and sharing your experience through copy. You’re also your experience of like, writing ads. Recently, I produced some ads and had my team produce ads on a project in the franchise space. And one of the things with franchisees is you get a lot of people who you get a lot of small bosses, you know, you get the franchise or the big boss, but then you have all these small bosses who buy into the franchise system and have an expectation that you listen to them. And we generated some audio for some ads and used Eleven Labs for it. So that’s your first AI tool 11 Labs great for audio voice. And I did it with a couple different voices and people like that voice doesn’t work for me. 

00:18:49 Casey: Like to me, I listened, they said I listened to that voice and I don’t like it for my community. That’s fine that you don’t like it. But my experience says that it’s not what I like or dislike. It’s a creative that we should test and I have experience stating that we once ran I once ran like two years ago, an ad with a woman’s voice that seemed atypical. It was a Jamaican woman’s voice in a predominantly Caucasian community. And the ad was so successful, it doubled results. It got us twice the leads for the same cost, or it got us leads at half the cost, however you want to it. The same leads at half the cost. Amazing outcome. My experience says that this is worthy of a trial. Everyone else lacks that experience. 

00:19:34 Casey: Their experience is, hey, Casey, we don’t have Jamaican women in our community. I get it, but I’ll tell you, this ad was a magic ad for us. So to me, it makes sense for us to try that. That’s my experience. AI wouldn’t do that for me. AI doesn’t have that taste, discernment and experience. So I want you to become a Renaissance man or woman. I want you to commit to learning the hard stuff, how things work. My toilet was running, we have like a powder room on the first floor and the toilet was running. I like heard it. And I was like, that’s weird. So I open up the back and I see the flapper and there’s like some scum on it like some like, what even would it be like mildew? It’s not mildew mold. I don’t know. It’s just like slimy. 

00:20:16 Casey: So I just like flush the toilet and I scrubbed that out. And I closed it and then the seal worked. Like there’s an experience that I know how a toilet flapper system works to flush. You need to develop those things like these like, understand these gaps that you have mentally before you outsource any of this work to LLMs.

You could turn around to an LLM and say, Hey, I’ve got a running toilet. Tell me how I should consider fixing it. Tell me how they work in the first place. How does, how did the valves work? How does the water stop? Like interrogate the LLM to understand that’s fine. 

00:20:53 Casey: But don’t say, tell me what to do next. Like first understand the problem in totality and then what to do. Here’s a fun example. So over the weekend, I’m recording this just after the Christmas kind of long weekend. I ordered an emergency power off button. It’s a big red button that’s on like a little plastic housing that you slap it and it turns power off like at a factory. I got on Amazon. I think I spent like 10 or 11 bucks on it. Pretty cheap. And I had this vision of what if my kids had this button and if they hit it, all the lights in our living room would turn off. We have some like globe lights that have uh hue light bulbs in them. And what if they pulse RGB colors like dance. And then what if our smart speaker plays Bluey’s dance mode? And if you’re a parent and you know Bluey, you know just how cool that is. 

00:21:47 Casey: So I thought through that is like, that’s so fun. And I went through step by step by step asking the LLM like, okay, like I was asking ChatGPT this, okay, I want to do this. Can I do it with this micro control that I already own? And the LLM was like, yeah, you can. All you have to do is short pin two to ground. Like that can be your button. Okay, cool. This is, I bought the thing from Amazon. I look at it, like, I don’t really know how this button works. There’s like a, it’s like an NO and NC circuits, which is like, one is like always closed. The other one’s always open until you like press the button and then it flips or something. I don’t, I don’t know. But there’s like four screws. I was thinking that there’d be two screws. 

00:22:24 Casey: So I was stuck and I asked the LLM, I was like, what do I do? Like, what do these things mean? Help me understand what that means. And I was like, oh, this means that this circuit is always open until you press it and press the button and then it closes the circuit. The other circuit is always closed until you press it and then it opens the circuit. Like, oh, that makes sense. And then through that, I said, here’s the wires that I think I want to put on these places. Is that right? It’s like, yes, but you got to flip them over. So I’m asking the question to interrogate what works and what doesn’t. I get it all built out. I flash on ESPHome, which is for Home Assistant, it’s like an open source firmware and got it to connect to Home Assistant and then was able to put in some automations. 

00:23:02 Casey: But even with the automation, it’s like, how do I think through an automation of the lights? It’s so simple. Like, and I tell you this because it has application marketing. I’m like, okay, I want the lights to turn off. And then I want them to pulse different colors randomly, but I want them to be in unison. Like I know that that’s what I want. So, okay, LLM, help me think through this. Like, is this the right way to think of it? They’re like, yes. And it programs the whole thing. It’s great. And I get it live and the kid hit, my son hits the button and the Bluey dance mode comes on and lights aren’t flashing. And was like, yay, we did it. 

00:23:30 Casey: And then they’re like, how do we stop it? It’s like, oh, I don’t know how to stop it. Like we unpress the button and it doesn’t stop it. The lights just keep flashing. It’s like, whoops, I didn’t consider that. So then I had a thing that I programmed it to stop it. And then I realized, well, actually, I want to go back to the original state. So the first thing I want to do before we press the button is I want, you know, when you press the button, I want it to record the current state of lights. And then I want it to do the dance mode.

00:23:52 Casey: And then when you unpress the button, I want it to revert back to the previous state. So these were like individual steps along the way. And this is such a fundamental difference than going on GitHub, grabbing code, loading it on a board and just praying that it works. This was me like vibe coding it and finding problems along the way. And you know, have another solution that I want to have, which is I want to personally make a soft button on my phone that is like dad’s override that can turn it all off. So the button no longer works. And maybe that’s gonna be important at some point. you know, my kids are just like, just keep hitting and it’s driving me crazy. So I just do the override button or something like that.

00:24:31 Casey: It’s like these little things of like understanding the uh logical steps along the way is the requirement. And now I get it. And AI could do it for me because I can prompt it more clearly. I can tell exactly what I want. I’ve thought through the microcontroller that I wanna use, all of that kind of stuff. Like it’s all kind of well thought out in my mind. And now AI can make it happen faster. Like that’s the consideration. AI can make it happen faster because I know what I want. So let me tell you about a recent time that I kind of phoned it in and I used AI and I’m not really proud of it. 

00:25:06 Casey: So I was working on wrapping a project and just life, you know, it was just life. I didn’t want to stay up late the night before I was delivering it. So I spent, you know, an hour and a half kind of working on some stuff to present for the delivery of a project. And then the following morning, I had a couple hours, I wanted to get a workout in, wanted to go for a walk, but I also needed to wrap this project. So I threw my notes into Claude and I asked Claude to come up with kind of a report for me. And it did. And it was okay. The report must have been 25 pages and so much of it was just like repetitive and annoying. 

00:25:41 Casey: And one level, it did shortcut the time for me to deliver a work product, but the work product that was delivered was actually reasonably low quality. So I just spent a lot of time editing it. Now, is there a net benefit to me using AI? I mean, I did it faster and with all the editing, I still got a work product done faster. I’m not all that proud of it. And it really felt like I cheated. It felt like when I was in school and I learned how to do long division and then I stuck the calculator out for a problem and I typed it in and I got an answer. That’s like, I just cheated myself a little bit.

00:26:18 Casey: You know, it’s like playing video games as a kid and having Game Genie on so that you have uh unlimited lives. It’s like, yeah, but is that really it? You know, it like, it doesn’t make you better. Yeah, you get the quote unquote, the dopamine rush of beating the game, but you kind of cheated. So I have this relationship with AI tools, which is like, I don’t want to use them for a lot of stuff. I’ll use them maybe a little bit to help with some of the statements into a proposal because that’s only gonna be a couple paragraphs and that’s easy to edit. And it’s fine for AI to take a first crack at it. That’s fine.

00:26:55 Casey: But man, I would say more and more, I’m not relying on AI for the first round of anything. If I haven’t done it myself before, I gotta go learn how to do it. If I’ve never written the ad copy before, I gotta go write the ad copy. If I grab a hook that I find that’s working really well on the internet, I, you know, that’s in my purview that I think is cool. And I rewrite that for the client and I write the body of the ad. And then I write the call to action.

00:27:17 Casey: And then I run that through Eleven labs to get uh an MP3 voice. And then I sent it over to a video editor to add it into an ad. Let’s say I go through that whole process. It is slow and I do one. I don’t love it. I do another one. I like it a bit more. Do a third one. I’m pretty happy. Do a fourth, fifth. I’m pretty thrilled. I think they’re good quality ads. They’re at my level of like where I want them to be. Now I have a system in a process. Then I turn to an AI and say, boom, rip me now 50 hooks. rewrite this body section five times, create these different permutations for me. And then I can tie that in with an n8n.io automation that could create these voiceover scripts for me if I wanted and get them recorded through Eleven Labs. Like that’s all possible. It’s cool to have that as a functionality, but first you got to do it yourself. 

00:28:08 Casey: You just have to do it yourself. Renaissance man, Renaissance woman, do it yourself. That’s it. The day, I see people join calls with me with their AI note takers. And just imagine that you can go get applaud, which looks really cool or a limitless.ai. I mean, I even vibe coded this shout ESP 32. Forget what variation this one is cents, s3 cents, ESP 32 s3 cents, it has the camera and microphone built into it. It’s got that I to see microphone. And I just vibe coded this, you know, one afternoon to do constant audio recording of me and it dumps that into uh a markdown file and boom, I can have like a corpus of everything I said or people said to me during the day that came out of my computer speakers or came out of my mouth. I can have all that content. 

00:28:58 Casey: What do I do with it? You know, it can summarize things for me. It can save me time. I think if I had like a learning disability, for sure, like that would be really useful. I think for me, for where I’m at though, I don’t use these AI note takers because I feel like it lets go of a skill set that I want to develop and hone and get better at, which is my own note taking. Because to me, note taking yields better memory, which yields better ability to solve problems quickly for the clients. I don’t want AI tools wrapped around me. 

00:29:34 Casey: Yeah, every single day I use ChatGPT. Yesterday, I was having a conversation with it. I was doing the dishes and it just had a conversation with

ChatGPT. And it’s super useful. When I get stuck on something, I want to understand how things work. I don’t want to go find a YouTube video that’s eight minutes long. I just want someone to tell me the specifics and I can drill down. I think it’s really useful. But to give away work to AI right now feels really dangerous. And again, there are individual parts that it’s appropriate for like content writing. 

00:30:04 Casey: Yes, but first write the content, you know what good content looks like, develop a rubric for how content is written, understand the different mechanisms inside the content that has to exist. And then, know, prompt your content writing AI. I use Claude right now for that. Like I’m sick of ChatGPT 5.2 copy. It just feels banal. It feels repetitive. It feels like everyone else. I like Claude for it prompted well, but that could change in a week. Just as the models update. So I’m using Claude. I’m using ChatGPT regularly daily. I’d say Claude now more than I have kind of ever since ChatGPT-5 came out. I’ve just been really unhappy with 5.0, 5.1 and 5.2. 

00:30:49 Casey: So those are the tools that I use. And what do I see our CMOs use? You know, by and large, there aren’t one size fits all tools. There’s the AI tools for LinkedIn outreach. I think that those can be okay. But man, if someone’s saying something in my voice, I don’t want it to be AI written. I want it to come from me in my voice. So if you’re gonna do it, you’d have to train these AIs on a big corpus of content. And if you are looking to building a vector database and attaching that with an MCP to Claude and having Claude be able to write stuff. If you don’t know what I just said, just Google it, mcpclaudepinecone.io. And just see if you can establish a Pinecone database for yourself, for your clients that can help you write content more effectively. It’s a fine way to go, but you really have to do some fine tuning. You can’t use these tools off the shelf.

00:31:37 Casey: You have to provide context. have to provide taste, discernment, and experience. You have to provide like a tome of your content that it can consume and understand who you are or who your client is. We call that core content. If you don’t have that, you gotta get that written. There’s a lot of work to do before AI can actually be all that useful. So am I long on AI? Am I worried about AI? I think the thing that AI has for us is speed but without taste, discernment and experience, that speed develops a work product that’s low quality. And I can’t bear that. I can’t bear to deliver my client a faster, lower quality outcome. 

00:32:17 Casey: Generally speaking, you know, if they’re like, if I had a client that called me and said, hey, this just happened, we need a new website up today, you know I’m using AI tools to make sure that happens. Leveraging the team to use AI tools for sure. But given the time and the appropriate heads up, we’re going to be building stuff the old fashioned way, and we’re going to leverage AI to help us a bit. Get us over writer’s block, fill in some of the blanks, do that kind of stuff. m But I think of like, photography, I’m into photography, it’s one of my hobbies. I’m into reportage photography. I really like Magnum, the photography agency. 

00:32:57 Casey: These are reportage photographers, started with Robert Capa, who shot the landing at D-Day, all those incredible photographs from, you know, the Allied forces on D-Day, just like storming the beaches of Normandy. That’s Robert Capa. He photographed him. He started up a Magnum photo agency with some other greats of his time, like Henry Cartier-Bresson. And Magnum still exists today. If you saw photos from the White House in mid-December of Trump’s staff, those were taken by Magnum photographer, Christopher Anderson, who is an incredible photographer, who’s got some great books about… Like he has a book called Sun and a book called Pia, that’s his son and his daughter. Just really beautiful work. 

00:33:40 Casey: There is a way that Magnum kind of has photos taken, like they follow the rules in such a beautiful way. If you go look at Magnum’s contact sheets or you go to the Magnum agency website and look at photographs, they’re powerful. It’s like a fundamental understanding of how powerful reportage, you know, like an investigative journalist style reporting looks. If you fundamentally understand that you can prompt AI to do it. But if you go into Photoshop, whatever it is, Photoshop 5 or 2025 or whatever it’s called, and you go use their generative fill, like, yeah, you can fix mistakes and extend backgrounds and do stuff like that. But it starts to lose its authenticity. 

00:34:26 Casey: And I think that there’s a knock on effect of losing authenticity that persists beyond the graphic that the team creates that uses AI. It like opens this door to speed at the cost of quality. And I want to tell you that you can live in a space of quality. Like the Renaissance man, the Renaissance woman appreciates the hand stitched rug. Like they understand the value of the uh hand carved wooden spoon over the plastic injection molded one because they understand the craftsmanship of it. And there’s just something about it that’s different. And I want you to have that taste and discernment and experience. 

00:35:11 Casey: And if you choose to use AI, you’re welcome to, but it all is predicated on your ability first to do the thing right the first time. And if you don’t know how to do it, you can ask AI for the shortcut, that’s fine. But also it’s hard to know that AI is giving you the right answer. Sometimes they give you the wrong answer and you know how frustrating that is.

00:35:30 Casey: I was wiring something a couple months ago on a little a little board. It’s called a whiz block mesh tastic starter kit. And I fried the board. I followed the instructions from ChatGPT. I didn’t think for myself and I fried the board. There’s like a $15 $20 mistake, but it was it just kind of pissed me off that like I let go of critical thoughts. People do that all the time with their marketing. They’re letting go of critical thought.

00:35:59 Casey: That’s where I want you to be, like understand it. Here’s another one. I’m working on a project and uh someone said like, I need the script added to the website. It’s like, okay, cool. I don’t do technician work, but that was just go to Google Tag Manager and add a new tag for all pages and just paste the script in. Took me from getting the email done two minutes maybe. So I ripped it and I did it. And then I walked away. And then I came back and it was like, script wasn’t posted. And there was like a JavaScript error.

00:36:29 Casey: And I looked at it and I was like, what? I went back to the email and I had copied it correctly. But the script that I was given in this email from the platform itself lacked the closing slash script element. And just like looking at it, I was like, okay, it opens with script, know, JavaScript. And then it doesn’t close with /script. So wrote that in there and saved it and it works now. Like there’s a sense of like experience there that I knew what to do, what to look for, and I could fix it. 

00:37:03 Casey: And I didn’t have to go to a developer to fix it. That’s what you’re looking for. It’s like develop in yourself this capability to understand these things so that you can get better. All right, I want to wrap this up and say AI I think is super cool. I think it’s important that you use it. I think that if you treat it like you treat a calculator, which is I no longer do math and I just use a calculator for simple math. You’re going to suffer. You’re going to turn into the type of person who doesn’t have critical thought, who loses a sense of logic, who doesn’t understand how the world works, who lives in a abstraction of reality. 

00:37:42 Casey: And I don’t think it’s a safe place to be. I think that’s what the billionaire class wants us to do. They want us to live this place, you know, disconnected from the world. I don’t want you to live there. I want you to live connected to the world, how things work, why they work. And then sure, use AI to get you there faster. For sure. I didn’t write all that code for my little dance mode button myself. No way. eh I wrote all of the lines, but I was able to prompt it intelligently because I knew the problems that I was trying to solve. Not just saying, hey, just like do it for me, right? With that travel website, I came up with the idea of let’s make it like a music festival. 

00:38:19 Casey: And then we looked at music festival sites and then we prompted with that. And then we used lovable.dev. And then later we used Bolt, B-O-L-T. That’s another great one for vibe coding a website. That those are really like the tools that I see the most utility in ChatGPT, Claude. I like Claude for its ability to connect to the MCPs, which is model context protocol. It’s kind of like an API to connect different things together. Look up Claude MCP. If you want to get lost in some really cool nerdy stuff around what’s possible. It’s very exciting. Anthropic developed the MCP protocol and what you’d call it, the MC protocol. They develop that. It’s kind of like an API again. So that’s cool.

00:39:01 Casey: And you would need cloud desktop for that. But those are really the places that I think that most marketing work is happening. It’s happening in ChatGPT, it’s happening in Claude. And then like front end web development mockups, like lovable.dev, bolt.new. Those are the two sites like that that I really like.

00:39:24 Casey: And then you can find em all these other like add on tools that you need. But for baseline AI tools, it’s really those four, really those three. I’d use Lovable, Claude and chat. That’s it. I don’t need to pay for uh lovable because I use it so infrequently, but it’s lovely to have. And I pay um just that $20 a month for ChatGPT and Claude. That’s it. I might upgrade at some point to a higher pay that ChatGPT just to get our team on it. There might be some utility in that, but at this point, not really. 

00:39:59 Casey: And if you want to go to the next level, just look up vector databases and pinecone.io as an example, that can add a lot more context to your prompting if you can bolt that into Claude and you can do that right now through their MCP. All right, I want to announce that we do have a webinar coming up. uh Depends on when you listen to this. So just go to cmox.co/webinar and sign up, it’s a free webinar. I’m excited for it. I’ve been working and building out slides for it. I’ve got some interesting and exciting stuff to share with you. So sign up for that. Again, it’s a free webinar. You get to spend time with me and then we do a Q and A session. So we’ll be doing that a little later in the month. So sign up right now to make sure that you can be on it. And I’ll talk to you soon. All right, let’s make this a great year. See you later, bye. 

00:40:44 Casey: Thank you for sticking around for the full episode. As you know, learners are earners, but you’ve got to take action on what you heard today. For more information and show notes, visit FractionalCMOShow.com. If you’d like me to answer your questions on an upcoming episode, you can share your question at FractionalCMOShow.com. And last, please hit the like and subscribe button so that I know that this content is helpful to you. All right, go get them.