AI Marketing Success

Speaker A: When I was thinking of designing things with AI, the first thing that came to my mind is, do I have a brand package that I could fit the AI and say, these are my colors, these are my fonts. This is the way that the header should look like. If you don't have that, then probably you're going to get generic content.

Speaker B: Hey, Guy Powell here. And get. Just getting ready to start the next episode of the backstory on getting an AI. And if you haven't already done so, please visit prorelevant.com and sign up for all of these episodes and podcasts. I, um, am no longer the author of the upcoming book. I am the author of the book. The book is now available, the AI Marketing Machine. And you can find out more about that@marketing machine.pro relevant.com or just go out to Amazon and buy the book now. It's on sale now both in a Kindle version and in a paperback version. Today though, uh, we've got, uh, somebody that I think really understands the challenges with AI and implementing AI, and that is Marvin Martinez of Bandsaw AI. He's the founder and CEO. And at Bandsaw, they help serve companies, uh, to implement practical AI systems that reduce manual work, save time, and improve efficiency without, uh, adding headcount. And, you know, that's. That right there is, you know, the good and the bad news. Certainly we're going to be able to implement more and do more, but it's really not only doing more, but doing more better. And that's pretty good. So, Marvin, welcome. So good to have you. Just, uh, look forward to our discussion.

Speaker C: Marketing's operating system has been replaced. Not upgraded, not improved. Replaced completely and totally replaced. AI isn't a tool you bolt onto your existing marketing organization. It's a fundamental redesign of how marketing thinks, operates and creates value. And most CMOs are already behind. The AI marketing machine is the field guide for marketing leaders who want to get ahead of that curve. Built on more than a decade of the Marketing Machine methodology and interviews with over 100 CMOs, this book shows you how to harness generative and analytic AI together without losing the human edge that makes great marketing unforgettable. This isn't theory. This is the playbook. How to redesign your team, how to build AI into the foundation of how you work, how to prove ROI in the boardroom, and how to lead the intelligent marketing machine that wins in the marketplace. Marketing's future is already here. The question is whether you're building it or watching someone else do it. The AI marketing machine available now, yes,

Speaker A: thank you so much for having me. I look forward to the discussion too.

Speaker B: Yeah, absolutely. So today, uh, we're going to focus on an interesting topic. And one of the challenges is, I mean AI has been implemented in Copilot and you know, with a lot of larger organizations and maybe even some smaller ones and doing some simple things, improving some email and maybe writing a couple of internal documents or even some emails that go out, uh, you know, en masse to uh, to, to our customers and things like that. But to really get the sense and value of AI, uh, we need to do kind of a, you know, a whole implementation, kind of like it used to be with an IT implementation. And unfortunately many of those AI marketing information, uh, implementations fail and we're going to find out today how to fix them. So let's talk about that. So what is the promise versus the reality of AI, uh, and marketing today? Uh, Marvin?

Speaker A: Well, you know, when you think of AI, you always think of how fast you can be in delivering the output. And think about this, uh, let's talk about social media for a moment. If you go to Facebook, uh, Instagram, even LinkedIn, when you look at a post, are you interested, able to identify that the post was written by AI in the first five seconds? I bet that the answer is yes. And the reason why that is is because AI promises you to deliver, uh, high quality outputs, but the quality of the output is going to be as good as the context that you feed the AI. So that's, that's one contrast. The promise of delivering fast and good content versus the context that you provide. Right? And when you start implementing AI, think about how you would train someone new. And this happened to me in my marketing strategy. Um, when I was thinking of designing things with AI, the first thing that came to my mind is do I have a brand package that I could fit the AI and say, these are my colors, these are my fonts, this is the way that the header should look like. If you don't have that, then probably you're going to get generic content. Right? So you have to be very careful as to what you feed the AI, the context that you provide so that you can really have like the output that is actually going to deliver high quality content to your audience. Because again, you can, you can see in the very first five seconds of a post if it was made by AI, uh, things like using the word most. If you see a post studying with most, like most companies, most people, that's AI written. If it has EM dashes, that's AI written if he has contrarian ideas. Like for instance, you know, uh, it's not about experience, it's not about effort, it's about discipline. When you have that contrarian ideas, that's AI. And you know, in this uh, era with so much people using AI for the same thing, the question in marketing is what makes you different? What makes you attract people? And that's when transparency come into place. And that's when you know, putting your face to the, you know, to the camera and record yourself and speaking, you know, out loud your ideas, that's what really makes the difference. And you know what my most uh, or my highest performing post. They are posts about me, what I do, who I am and my battles of my everyday. You see. So that's kind of the difference between what AI can provide and what it promises versus the actual delivery.

Speaker B: Yeah. How true that is. And uh, you know, it's so true. The uh, so the M dash is uh, they drive me nuts. And, and then um, and uh, one of the things too with Claude I m used to use uh, ChatGPT, but I use Claud now and Claude uses English English, not American English. And so the Z versus the S in some instances and, and uh, you know, and that, that right there is kind of a, a good giveaway. And then what you're talking about is, I think they're called the, the, the parallel negativism where you do the, you know, it's this, this, this and this. But it's not that, that and that or what, however that works. And uh, it's funny now, I hadn't heard the one about the most. I'll have to start looking at that. That's a pretty good giveaway. Um, now that though talks about the quality of the outputs and certainly you want to have especially anything that's customer facing. You want to have um, good quality. But what are some of the other root causes of uh, implementation failures that marketers maybe never even see?

Speaker A: You know, that's good because you have to really think through uh, your strategy. And when I speak with business owners about how they want to implement AI in their day to day, uh, specifically talking about marketing, sometimes owners don't have a structured plan. Like for instance I have this client who said I want to create social media posts using my website content. Which is fine, that's a great idea. But my questions like right away were what type of content do you want? How does the content look in your marketing funnel? Is it the top of the funnel? Is it the middle of the funnel? Can, uh, you tell me more about your voice? Can you tell me about your experience and your point of view regarding this topic? It's not only saying write a post about this topic, topic, that's not what it is because otherwise you're going to get a generic host. So most, uh, of the failures I've seen is people not really thinking through the strategy and not having their processes, uh, built out in SOPs or even in a piece of paper. You don't have to be fancy for that and that's what makes it fail. So this guy, he said, I don't have time for this, just do the automation. I want, um, my AI agent creating posts from my website, from my podcast, and that's fine, I built it for him. But guess what? He was not getting impressions, he was not getting comments, he was not getting leads from that content. And the reason why that is is because he was just purely genetic, right about one specific content, but that really like, didn't have a voice. So you mentioned using cloud, cloud as a tool is really powerful. So, and what makes it, and what is going to make it successful for you is how you fit again, the context. So I am a marketing strategy, uh, strategist. Uh, this is the way I do things. This is the way I think about the funnel. You give a readme file to Claude so that it reads every time before it does something and then you're going to become more successful, successful, more successful and more successful. You tell Claude, hey, I really like that output. Can you please save it in your memory so that next time you make a reference to that? Like, for instance, like one time I made a post, uh, with Claude that hit like 45,000 impressions in LinkedIn. And it was just about a general topic, uh, like something about the, uh, AI domain. It was very interesting, a lot of people commenting, a lot of people liking it, so on and so forth. And I used that as a reference and told Claude, hey, look, this was the impression count, this was the, uh, engagement of the post. Let's do more posts like this. And then that's when you increase your reach and your visibility for people. So you need to have that strategy in place if you really, you, if you really want to be successful using AI.

Speaker B: Yeah, that is so true. And um, you know, to your point, uh, you know, a, a social media post especially, you know, what you want to do is you want to get people to read it. I mean, what good is it, you know, if it just talks about, you know, something that you find interesting as opposed to your readership will find interesting and uh, and, and really understanding, you know, make, make sure that the post is something people are going to want to read about. And then of course, you know, there's other things that go into it and I'm not an expert on it, but, you know, what's the hook, you know, you have. I always look at it as the first three seconds buys the next 30 seconds buys the next five minutes. So that first three seconds, you know, you got to buy them to read the next 30. And the 30 seconds has got to be good enough to then get them to, you know, to dig deeper and dig deeper. And uh, and that's hard. That is really hard. That's where social media marketers really earn their, you know, earn their stripes, if they can figure that out. So they can, you know, drive up the engagement, the readership and the viewership and uh, what have you, you know,

Speaker A: now that you mentioned it, I truly believe that when uh, the content is written, it really needs to talk to your icp, which is another piece of like making the mistakes, right? Like who are you talking to? Because at the end of the day, what's the purpose of marketing? Getting leads, getting market, like marketing qualified leads, people that, you know, uh, start getting interested in your service. So how do you write in such a way that it always talks to your ICP in a way that if, if my ICP is a business owner, uh, from cleaning companies, you know, I want them to read my, my uh, social media post and say, that's me. You know, this guy is talking about me. That's what's going to make a difference. And again, important, you have to define your ICP first. You cannot create a, uh, powerful and impactful post for someone if you don't even know who that someone is. And surprisingly I have spoken with several business owners that they don't know, you know, uh, who their ICP is and they cannot define it. Apparently they know. But how do you define it? That's the issue because when you talk to the AI or when you make the AI agents or whatever, automation, you have to explain it. You have to tell the AI, this is what I'm looking for, this is the people that I'm targeting. And that's, you know, when most implementations fail.

Speaker B: Yeah, absolutely. And um, you know that one of the things from my book, the AI Marketing Machine, by the way, it's available now. Just kidding. But it is available now on Amazon. But in any case, uh, that one was one of the big findings is uh, you know, if you don't give the right kind of information to the AI when you're writing, uh, whatever kind of content it is, uh, you know, then you're going to get a generic answer. And what's really critical is the difference between generic answers and specific answers. You want to answer from your prompt. That's going to give you something very specific that's going to help you to, you know, in, in our discussion here, increase viewership, increase readership, get people past the three seconds, the 30 seconds and the five minutes. So that's what's critical. And one of the big pieces of that is, you know, really feeding the AI, uh, consumer insights. And uh, you know, so one of the things, you know, we, we need to talk about is, you know, so how does, how do, how do we handle community, consumer insights, uh, you know, to make sure that the AI really gives what we want.

Speaker A: Yeah, it's all about that uh, feedback cycle. Picture this. You have a, uh, lead magnet that you sent out to people via DMs, via uh, email list, uh, whatever, right? So, so how many people did actually sign up for that lead magnet? How many emails did you get? How was the reaction of people similarly, uh, with a social media post, how many likes did you get? How many comments did you get? How many shares did you get? And then you get all that information and you pass it on to the AI and say, hey, uh, you know, I had these many reactions, I had this many DMS from the post or from my lead magnet, right? So I just want you to save that information and use it for the future and you start building your own strategy and you start seeing, hey, this type of content really work. It's like AB testing. So you do one post, you see the reactions, you see, uh, how well it performed, you feed the AI back to it and then you can have the AI analyze the comments. We have tools and that can connect to your social media, like ManyChat, even Cloud Code, the extension, you can put it into your, uh, Facebook account, Instagram, you can have it read all the comments and see how people react to that. And then from the comments you start creating more content and more content and more content. So it's really important that you feed the AI with the actual comments, likes, reaction from people that is interacting with your post. And I can guarantee you that you're going to get better outputs every time. And every time.

Speaker B: Yeah, how true, how true. You know, and one of the things too is, you know that consumer insights or customer insights is uh, is so critical and um, one of the things that is, uh, you know, potentially a way to, to improve your AI is really understanding how your customers actually think. Uh, tell us a little bit about that.

Speaker A: You know, that's, that's a tricky one because it only, it comes with repetition. It comes with you doing the same thing over and over again, maybe different formats to see what actually works. I, uh, I've been working with a team that has tried five different strategies to see what really resonates with the public. But there's one principle you have to think about what pain point you are solving for. Support other people. So if I'm thinking about business owners, right, what problems do they typically have? Like, do they want to increase their sales? Do they want to improve their sales? Uh, strategy, do they want to get, uh, better at customer service? Like, and then the content goes again, uh, targeting those pain points. So it's, it's really important that those signals that you see, uh, you know, you use it to your advantage. If you go to Reddit, for instance, I have an automation that goes to Reddit to specific groups that I typically participate in and see what people are saying, hey, you know, uh, I have not been successful with this. I am having issues with that. How do I, uh, get better at customer service? How do I get more Google, uh, star reviews? And then from that content, you start thinking and making your strategy. By the way, a strategy and critical thinking AI cannot do. That's you doing it. It's you driving the car. AI is the car, you have to drive it. Right. So all those insights from the customer, that's one way of getting it going to, uh, as I mentioned, Reddit, uh, like Facebook, like, whatever, right? And then just getting those insights, like putting into the AI and say this, this is what people are saying. How do I match that with my offer? How do we make it attractive? But it's all about you speaking to the AI, thinking just like your customers would think, and then start it from there.

Speaker B: Yeah, absolutely. That makes so much sense. So, um, yeah, tell us, maybe you could summarize a little bit. What do you see as the real roadmap to make sure that AI and marketing, that marketers just get it right? What do you see as the top four or five pieces of that roadmap?

Speaker A: Yeah, the first one is make it right manually. If you put AI in anything, it's just gonna amplify what you have. So what that means is if your content or marketing strategy does not work at all, don't start thinking about how you can automate because if you do that then uh, the, the wrong output is going to be multiplied and it's going to be multiplied by m by a lot. Right? So that's, that's the first thing in the roadmap. Put your things together, see what works. And then when, when you see something working, then start thinking about automating it. Think about workflow. Step number two, it's not only about, you know, these words. For me it works. It's also about how do I put it in a paper in such a way that it's easy to follow. So for instance, how do I do my research to see what type of content I'm going to do? How do I transform that research into actionable items? How do I get those actionable items into resources I can share with people to add value? How do I collect uh, information from people that is reading my content? And then you start mapping the process. When you understand your workflow as the second step, adding AI to it is going to be super powerful. I have worked with companies, uh, where their automation is only four steps, only four steps. But because they thought through it, they tested it, they know What a step 1, 2 and 3 and 4 looks like. It's so repeatable that they do so much with so little. Right. So having your workflow as the second step in the framework is super important. Number three, define your tools. With so many AI things out there, it's easy for you to distract. You start looking at club code and all of a sudden you have anti gravity doing some cool stuff. So you switch to anti gravity and then as you are working there you see Codex, uh, doing some fancy um, like tools or some fancy results action features that you may say, hey, this can work for me. And then you go back to cloud because cloud got improved or got upgraded. So when you start thinking about how many tools that are out there in the market, you have to specialize in something. Don't try to use many tools at the same, like pick one tool and get really good at it. My favorite tool is for automation is cloud code. I agree with you 100% plus N8N when I have to connect multiple sources and multiple softwares uh, altogether so I use that, uh, and lastly fail fast and fix soon. What am I saying you don't have as the last step, when you find something that works, you put it in a workflow. You use your tools. Don't think about complex uh, projects, just start doing right. So if it's a one week implementation, just do it, start implementing. See the Output and then fix it. Because I guarantee you, it doesn't matter how much you plan for it, you will always have edge cases, you will always have an output that you will not like. And no one will make the perfect output in the first try. So if you have to iterate many times before getting the actual output that you like, then start soon, fail fast. Do faces in your project. Phase one, phase two, phase three, right? Try with the first one, see what works, see what didn't, see if it was helpful or not. And then move on, see what adds value. Because at the end of the day when you start thinking of complex and long implementations, man, people get tired, you don't see results and you know, it all gets frustrated. And then companies say AI is not for me. And that's, and that's a huge mistake because AI is part of the day to day and you have to use it if you want to, to stay competitive.

Speaker B: Yeah, yeah, that is so true. And uh, because your competitors are using it, you have to assume that they're using it and they are either, you know, they're learning just as fast as everybody else is and if you're not going to keep up, then you are going to fall behind. It's such a critical, uh, you know, way to think about that. Um, this has been awesome. Uh, Marvin, I do have, excuse me, I do have one last question for you. And uh, that is, what advice would you give an up and coming new marketer that's trying to break into marketing when he's uh, when he or she are potentially competing against, uh, an AI chatbot?

Speaker A: Well, I would say, uh, start thinking about what makes you unique that you can bring to the table. Because everyone can use AI. Like everyone. And the tools are everywhere. You will have free resources, free tools, YouTube videos, but uh, like I mentioned earlier, your creativity, your experience, uh, and everything that you have learned personally, your personality. In fact, when you add that to the AI, that really makes you powerful, different and unique. So think about what's uh, your unfair advantage. Taking into account your own experience, taking into account your own skills, personality, what you like to do, topics that you like to talk about, be passionate about what you're doing and you'll be successful.

Speaker B: Yeah, I, I couldn't agree more. And uh, and actually while you were talking before, I was taking notes and, and that right there is so critical for a young marketer. But uh, everything, especially your roadmap makes, makes so much sense. Um, you know, so Marvin, I wish we had more time. Maybe we'll get on in another couple weeks or months and uh, and do the, the next update because there are so many new things coming out and this industry is changing so fast. Uh, where can folks learn more about you and your company?

Speaker A: They can go to my website, which is Bansaw AI. You, uh, can actually look at it or you can just chat with my chatbot. Of course. I have a chatbot right there that is trained with the way I will respond. So if you have any questions, just put them in. I get them in the background. Or you can connect with me via LinkedIn. Just go to the search bar, put Marvin Martinez, and then I'm going to pop up in the third or fourth result, hopefully, if not the first one.

Speaker B: Absolutely. No, you're going to pop up in the first one, that's for sure. Well, let's make sure that we connect as well and go, uh, from there. Uh, thank you so much. Uh, this has been really good and perfect and it's such an important topic because it is so easy to get the implementation of AI, uh, and marketing wrong. And getting it right can have an enormous impact on sales and your brand and even your career. So, uh, very important topic in any case then for the audience, definitely. Stay tuned for many other episodes in this series on the backstory of marketing and AI. And as I said, I am no longer the up and coming author of the new book the AI Marketing Machine. I am the author of the book now. Ready for sale, the AI marketing machine. Go to Amazon.com and you can download it and go, uh, from there. Marvin, thank you so much. I really appreciate it.

Speaker A: Thank you. Uh, guy, it's been a pleasure.

Speaker B: Absolutely. Thank you,

Speaker A: Sam.