The AI Marketer's Playbook
The AI Marketer's Playbook is an actionable podcast focusing on AI and marketing. Each episode covers AI strategies, tools, and trends that are changing marketing. Listen to interviews with industry experts, analyze case studies, and get practical tips. This podcast is for anyone looking to leverage AI in marketing to improve results.
Tune in to stay updated on AI marketing.
The AI Marketer's Playbook
58 | Kuba Czubajewski on Building Human-AI Content Systems
What happens when you blend human creativity with AI precision? Kuba Czubajewski, founder of Story Angle, reveals the system behind scalable content that doesn’t sound robotic. In this episode, Audrey Chair dives deep with Kuba into B2B content strategy, positioning, and the real reason most AI-generated posts fall flat.
Discover how to use AI assistants (not agents), craft LinkedIn posts that don’t sound like everyone else, and why founders must do audience research before writing a word.
Join my weekly Newsletter: https://lp.closewithcopy.co/welcome
Welcome back to the AI Marketers Playbook, where we cover actionable frameworks to help you leverage AI and marketing strategies in your business. I'm Audrey Chair, your host, and today I have with me Kuba Czubajewski, the founder of Story Angle. Now, Kuba has spent the last seven years in B2B content marketing working with. Over 40 companies and now building a content led growth system for B2B founders using a blend of AI and human methodology, Kuba has developed a method for building highly personalized AI content creation systems for Lean B2B teams, and has. Dozens of companies embed these assistant directly into their daily operations. Today, we'll dive into how you can use custom AI assistance to supercharge your content workflows and what it really takes to train them. Well, Kuba, we are excited to have you on the show.
Kuba Czubajewski:I'm excited to be here. Thank you so much for the invitation.
Audrey Chia:Awesome. Now, UBA, tell us a bit more about yourself, your background, how do you get started in perhaps marketing, and then when do you start integrating AI into your workflows?
Kuba Czubajewski:Yeah. So, um, I started in marketing. Uh, in, in a very unpredictable way, uh, because it was a complete accident. Uh, I found, um, I actually found a Facebook ad about content marketing course, some kind of content marketing course, and I got interested in that so much that I started. To learn it and, and study it properly. I come from a very creative background. I grew up in theater. I was a child actor my whole life. So, uh, and I was after, after I finished high school, I wanted to do something creative with my life and turned out content marketing is some, is something just like that. And since then I've been developing my skills. I've been learning so much about content and started my, my own agency, which is story angled. And with ai, AI is of course pretty new overall, but I started with it pretty quickly. I jumped on, on the AI train, uh, pretty fast. I wanted to see how it can, IM improve the, the current operations there. Content, creative process. And from the very beginning I was using it as an, as a form of expanding my own thinking and expanding my own, uh, my own processes. Uh, as, and so far it's, it's been, uh, a very, very great journey.
Audrey Chia:Wow. I love that. Even for me. After integrating ai, it's really something that has supercharged my own business. I'm sure it has done the same for you and your clients. Now, tell us a bit more about B2B content marketing.'cause most people are familiar with marketing, but they don't really know how this applies in the B2B context. What is the difference between, you know, B2B and B2C marketing?
Kuba Czubajewski:Yeah. Uh, great question. Um, the main difference, uh, I would say is that in B2B, you're not selling to just one person B2C. Content marketing is, um, simpler in that sense. Not very simple, but simpler in a sense that you have, you can create a very specific persona. Target the content toward it, and uh, and that's it. With B2B, you are selling to a buying committee, so people who, a couple of people in within the company, uh, decision makers, influencers, people who, who send stuff to their managers to, to check for, for them to check out. Uh, for the, the people who will be using the solutions that you provide. So, uh, the, for me, the biggest, the biggest challenge is to develop a B2B content strategy that, uh, targets all of those personas and all of those people within the company. Uh, while maintaining a unified, you know, image of the brand and, uh, making sure that the, the messaging isn't all over the place. And that's also why I, uh, got excited about, um, B2B content. That's why I, I spent most of my life, uh, professional life. Learning it and, and, uh, helping my clients with it because it's, um, it's a bigger challenge. It requires you to dive deeper into the psychology of the, of the people that you will be seeing your content. Uh, you have to juggle a couple of more things, uh, for, for it to work properly. So, uh, I say, I would say that. For me, B2B content is the, um, the way that you can talk to not just one person, but but a whole company in a compelling, uh, compelling way for them to trust you and invest sometimes a lot of money into your product.
Audrey Chia:Yes, and I think it's also a longer sales cycle where you really have to take time to nurture them across multiple touch points versus a B2C customer who can just make a decision and click add to cart. Whereas in your B2B space, there's so many stakeholders to juggle, right? So maybe I can also ask what would a typical process be like when it comes to that human approach? Then later we can talk about how you start thinking about the AI part of things.
Kuba Czubajewski:Mm-hmm. Yeah. The human approach starts with a very, um, comprehensive process of understanding the, the audience and what you, what they're, uh, what they're actually thinking. So, uh, there is this. There is this conviction in B2B content that, you know, you, you can create one persona, you can, uh, you can just, um, apply demographics, psychographics, some, some simple things and just create content based on that. From my practice and from, from what I've learned, it's not really how it works and the, the most important thing for me while building the, the human part of the content strategy is identifying those micro moments when people. Uh, when people start thinking that they might need a solution or they, they're experiencing a very specific problem in their, in their professional life. So, uh, the way that I approach creating the strategy for, for my clients is to first uncover the. Deeper needs, deeper goals, frustrations, aspirations, things like that. Um, this comes from, you know, professional stalking, honestly, which means, uh, which means just roaming around, uh, LinkedIn, Twitter, Reddit, and reading through. Thousands of comments and, and, uh, and posts to identify what actually makes people annoyed about their work or what actually makes people worried or what keeps them up at night, you know, things like that. And then. The second part of it, once we have all the, all the, uh, things gathered, the second part of it is to build, uh, you need to build a positioning, uh, of your product or your service in a way that immediately communicates that you're targeting at least. A portion of those things that you discovered? So, uh, a big part of my work day-to-day is helping my clients clarify what they actually make happen for their customers. Not necessarily what they do, but what they make happen in their lives so that they can feel like, okay, I need to talk to this, these guys because, uh, I feel like that that is my, what, what I might need, uh, in the current moment.
Audrey Chia:I actually really enjoyed what you shared about positioning, right? I think a lot of people don't realize the importance of it, but for those listeners who may not know what it is, it's really in simple terms, how you stand out in a crowded marketplace. So when somebody thinks of your product, they need to know why should I say yes to even, uh. Looking, a website hopping on a call. Right. And for many clients, I'm sure you might have noticed their products can do a hundred different things and you have to figure out what is their right, you know, positioning for them. How do you navigate that with your clients? Mm-hmm.
Kuba Czubajewski:Um, the main thing, um, is prioritizing the. The pains and prioritizing the, um, the biggest transformation. It's, it's very individual, honestly. Uh, from, from product to prior, I've worked with, usually I work with more technical B2B companies. Um, when the, to give you an example, my, uh, my client from the, um, from the. Business intelligence space, he creates, um, you know, business intelligence workflows and dashboards and things like that for, for those companies, uh, for manufacturing companies. And those companies have, um, they, they need a very straightforward way of, uh, things. Being named about what, what David, my client does, because this space is so complex, uh, and many, many other competitors are just, you know, overcomplicating it already with their, with their copy on their website, with their marketing messaging, with their posts. You know, it's all technical jargon and decision makers in those company, in those companies that David works with. They don't care about that jargon. They don't care about, you know, knowing all the fancy terms. So we needed to really simplify what actually David does for them and what actually help, he helps them prevent what, you know, what tragedies in their companies. He helps them, he helps them avoid. Uh, so that's one thing. The other thing, the other client, for example, who is a SaaS MVP development company. Uh, they focus, we focus their marketing more on the dreams and aspirations side, and we position them around building the product, um, the, the, the dream product of their clients without all the hustle that comes from working with development agencies, you know, and development partners. It's, um, it's so, as you can see, you have to first identify what's not working. Usually in the, in your competitive landscape. So what other companies do that don't really work and the easiest way is to propose something that works around that or, or kind of turn, turns it around and shows, shows your, your potential client who has talked to. Tens at least of companies, uh, like yours, uh, that you are doing things a little bit differently, that you're doing things in a way that is simpler, more accessible, uh, and especially with busy B2B audiences, this approach works really well because they're short on time. They got plenty to do. They got, you know, uh, even though the, the sales cycle is long, uh, and the decision printmaking process is long. It's much easier to break through when you have a very clear way of telling them, Hey, I'm, I will not put you in a position that this previous company tried to put you. You know, so, so I think that's the, that's the main thing.
Audrey Chia:Yeah, I, I definitely resonate with that, especially what you said about the competitors, right? So what are they not solving that you can solve? I, I like to tell my clients, it's like a triangle. Your competitors, your audience, and then yourself. So you need to find like that sweet spot between all three, and that's that, that real unlock. So tell us more maybe about the AI part of things. How does AI fit into this whole picture? Assuming that you need to tailor content to so many different stakeholders, where do companies even begin?
Kuba Czubajewski:Yeah, so, uh, the main, once you have all that knowledge about the audience, once you have all that intel, uh, it's much easier to start building, uh, AI content. So. You cannot, uh, there are, there are no shortcuts to this. Even though AI seems like a fast solution to a, uh, to, to, to the content problem. It is not you have to do your homework first. You have to learn a lot about your audience. And then the way I do it is I create, um. Also, maybe before I tell you, uh, one more, one more thing that we do before, before building the agent is to actually, is a short workshop where I try to understand the language that my clients wants to use. The, the, the tone of voice. We analyze the competitive landscape as, as well to understand how. How people, um, are reacting to different types of vo voices and, and, and, uh, approaches to content. So once you have that, you can start building shaping the, the content agents, which, uh, should be very specific, should have very specific roles. So, uh, for our clients, we built a separate agents for LinkedIn content writing. Editing, uh, LinkedIn, infographics, um, carousels, um, blog copy, uh, website copy. You know, all of those things are very, very. Specific roles that you, you, and you train your agent to, to know, and to, to recreate and amplify with, with your own knowledge. Uh, so, um, that's, that's my, that's what is behind this human, human plus ai, uh, methodology. So, uh, a deep, deep knowledge about your customer, deep knowledge about their habits, about the, the niche. Amplified by, uh, custom trained ai, AI agents or assistants. Um. With a very strong human supervision. So we do not automate things aimlessly or, you know, just, uh, just because we can, uh, we still need to have a very tight, um, control over, over the process to make sure that the content meets the quality. Uh, it makes things faster, it makes things more efficient, uh, but it still needs lots of human input and this will probably stay like that forever. The longer I observe this space, I believe this will stay like that forever because, um, because good content. Doesn't work in patterns, it doesn't work in, um, in, you know, repeatable. Um, always the same templates that AI uses and AI knows because it's, you know, it's a predictive model. It's, it's a statist statistical model, so it uses, you know, things that, that probably will be the best option here. And content doesn't work like that. Content is, sometimes it's raw, sometimes it's unpredictable. Sometimes it, it has to be very honest. Uh, for it to work properly and to resonate with people who are tired of all the repeatable AI sl that we see on social media lately. Uh, so that's why I think this methodology is the sweet spot because you have. Something amplified it, it makes things easier. Uh, but at the same time, it keeps the human element of it so we can, um, so we can keep the content original and feel like our client speaks to, directly to the ear of their, of the stakeholder or the, the target audience.
Audrey Chia:Definitely. I think it's also interesting that you mentioned how ai. 10 certain patterns, um, and it follows certain systems. I think actually if you are a copywriter, or at least you take a deeper dive at AI's copy, you would tend to see repeatable structures even in the same, um. Long form, which, which then makes you feel like, hmm, I think I just read this a few paragraphs ago. Uh, I noticed, you know, this mini quote that chat GPT particularly has, I think it might get better, but it's true, right? Because if everybody's using AI as a, to create content, and if they're not doing that, adding that human touch, then that's where it can seem very flat across the board. Now, Cova, tell us also about that. AI agents that you're talking about. So you mentioned briefly on agents and assistants. Are you differentiating them? How does it work in your workflow?
Kuba Czubajewski:Um, I call most of the things that I create assistance, not agents, because agents from the definition, they should be autonomous and they should have, you know, deep processes, um, integrated into them. So it works, uh, much, much more like a, like an employee rather than just, you know, a bot and that you ask specific questions to and, and response. I have a problem of the way that. The two terms get mixed up together in the marketing space, in the AI space, lots of creators, uh, call agents, which are things that are not agents, which are just, you know, you, you ask a question and it responds it. This is not an agent. Uh, and, uh, so I call them assistants mostly because it's, uh, at every step of the process. Some, some of the parts are automated, but at the same time, at every step of the process, the most meaningful steps, uh, are overview by humans. So. We are far away from, or probably we'll never reach that point where you can just, you know, type in to an agent, create me a week of content, uh, and it'll create a whole week of content, of impeccable, ready to post content. Uh, correct. I bet someone who, uh, doesn't agree with me to, to challenge me on that, but I believe that. We won't probably ever reach that point. And that's why I call my things that I create assistance.
Audrey Chia:It's interesting you say that,'cause a lot of people, they mix the two things out, right? Agents and assistance. So I personally have tried agents, I've also tried assistance. But I think like what you mentioned from a copywriter's point of view, sometimes the output that the agents give you is really flat and you're like, Hmm, I, I thought I specified the steps. Uh, you know, to get there. But actually, if you use assistance in which you can control the steps a lot more, and then you can have that human in the loop a lot more, you tend to get higher quality output. Have you also noticed, like when you use assistance, um, do they. Typically give you the same type of output, and if so, how do you make it unique to each client?
Kuba Czubajewski:Um, so, um, there are two parts to this. Uh, first of all, uh, the assistance for, for each client, the assistant, um, is, um, taught because it's prompted to, to do it that way, to find every time it creates content, it should find fresh angles for, for the content. There is a, you know, a huge prompt. Specifically for that. So, uh, so it doesn't fall into this trap of recreating the same content again and again that it sees from in the knowledge base of the assistant. Uh, so we, we work around that using, using this. Uh, and for, for each of my clients or maybe across my clients, the, this problem doesn't seem to exist because, uh, each assistant has. A quite a different prompt, so we don't use like, uh. You know, the same prompt template for all my clients. Each prompt is highly personalized. It has a separate knowledge base. Each assistant has a separate knowledge base. Uh, it uses separate tools to extract the data from the vector stores and things like that. So. So we want to, uh, we want each our of our assistants to feel like a companion that just joined the company of my client and is learning more and more about what they do is learning more and more, uh, of their style of content. And, uh, this helps me avoid what you just said. Uh, the, the, the, the fact that, um, you know, just a plug and play assistant that, uh, recreates the same kind of things again, it gets boring, it gets annoying, and, and you give up on it. Yeah.
Audrey Chia:Yeah. I love that. Could you perhaps share a demo with us? I'm sure a lot of the listeners would be interested to see what this actually looks like and how it actually comes to life.
Kuba Czubajewski:Of course. Uh, let me share my screen. Okay. So, uh, this is the app. Uh, maybe let me make my screen a little bit, uh, smaller so we can see it better. This is the app that we use for all of our, um, all of our assistants and, uh, here are, they are called agents. They're not really agents, uh, but that's another topic. Um. As you can see, uh, the, the three letters or two letters, um, uh, abbreviations here, it's, it's the codes for my clients. Each client has a different, um, different assistance for LinkedIn editing, ring LinkedIn, uh, LinkedIn writing. I also have my own assistant, uh, for newsletters, et cetera, et cetera. So, uh, let me open one of them. So as you can see, we have, uh, here. A long prompt, very long prompt. Uh, so it has, you know, a lot of, a lot of things to avoid, a lot of things. The, the structure of the post, some examples, a fabulous example and a weak example of content. Um, what, what is the depth of the content? So how deep we should get in, in, in what we do. For example, the more technical clients need a little bit more depth in the content while maintaining it. Jargon list. So we have to level it out. It takes a little bit of testing tone, a voice writing style, um, about me and my business, uh, section and what, what the, what the assistant actually does. And then, uh, the, uh, the knowledge section here, uh, is, uh, where I add all the, all the knowledge about the, uh, my previous posts, things that I wrote, things that, uh, that, uh, I publish on my LinkedIn, uh, and, uh. Uh, each of my client has a dedicated space within this software, uh, that, uh, includes their past posts. Uh, let me share this, uh, with you. Mm, their, their pass posts, their, their, their website, that this tool can also scrape your blog. Um, and it, it's much easier for, for you to, uh, to understand, uh, to, to kind of gather everything together. So if you open my, my folder here. As you can see, uh, this is, this is a sample, uh, that I usually use on, uh, um, when showing things to clients. Uh, as you can see here, I have a, an example of my, of my posts and post statistics in the CSV format to help the, the agent understand what posts perform better, what posts perform worse, and, and things like that. Um. So, and, uh, all of my clients have the same kind of folder. It's much more extensive, uh, with, with much more data points. Uh, and it helps me, it helps the agents. Um, each time the, the agent starts working or the assistant starts working, the, it scrapes all the data, it understands it, and, and then starts to write posts to make sure that, uh, it stays within the same realm that we've established, uh, for this particular client.
Audrey Chia:Wow, super powerful. And for listeners who may not be able to see the screen, uh, is there a name for this platform or is it something that you've built separately?
Kuba Czubajewski:Uh, this platform is called Dust. Dust, uh, tt, uh, it's the, it's the website for it. Uh, I'm not associated with dust in any way. I just like them. I, I like how they, how they do things, uh, and it's much easier for me to build those assistance within this software. I tried, uh, building custom GPTs, uh, custom gpt are difficult to, uh, to personalize as much because, uh, they, they can't, you can't have the system prompt changed, you know, so you, you cannot change actually what the, what the, the, the kind of overarching, uh, thing that chat GPT does itself. At one point you could do it using the OpenAI API platform. Uh, but they recently decided to, uh, to sunset the, the assistance, uh, API, and they, they are using the, a different form of things, which is more technical, more complex to build. And I like dust because I can share this with a client, they can use it themselves. And you know, we all stay within the same. Software, uh, it's all private, it's all, uh, it's all specifically for them. So it's much easier for me to work with multiple clients using this software rather than, you know, built customer assistance for each of them.
Audrey Chia:Yeah, I love that. Something, something I realized when using chat GPT custom assistance, it's, it gets lazy. I don't know if you have noticed that, but after like the fifth response, it gets really lazy.
Kuba Czubajewski:True, true. Uh, and a good way to, uh, and this is something that does also does, um, the, when, when things get, when the assistant gets lazy, I switch between while, while writing content, I switch between the content writer and the content editor. So, um, let me maybe share my screen again to kind of show you what I mean. That'd
Audrey Chia:be really cool.
Kuba Czubajewski:Okay. So for example, um. For, here's a content, uh, content that I wrote, uh, today for, for one of my clients. So I asked it first to, uh, to uh, just write, uh, a couple of ideas for the content. Uh, then I asked it to, uh, to write a couple of posts and. First I specified what the post should be about, and then I asked the editor to give me a brutal feedback to the post above. Uh, so the switching between and the data that AI gives itself from those two assistants. Kind of prevents the, uh, AI from getting lazy because the editor is designed to always push, always push for more, always push for better. And this helps my, uh, the, the, the writer not to recreate the same type of post again and again, the same structure and things like that. So, as you can see here, the, uh, no fluff feedback from the editor is quite extensive. It gives you a recommendation what's, uh, the weak points, the, the strong points. Uh, and, uh, asks you to rewrite the post using the, the recommendations, and then the, the editor rewrites it. I ask the writer to write another post and the process repeats itself.
Audrey Chia:That is so smart. I did not think of that. And that's a very. Brilliant move to get AI to critique itself, especially when you build two assistants for that. If, if you guys haven't tested, um, you know, building custom assistance for writing, I would like to highly recommend that you have a keen eye on your work because AI gets a little bit dizzy and then you would see. Getting a bit more slip shot with its work, and then sometimes it produces content that goes off tangent. So that is something I've noticed. And then that's also why I think, like what you said, the human in the loop is so important, right?
Kuba Czubajewski:True. True. You, you cannot, you cannot shortcut it. Unfortunately, uh, content is still a human job.
Audrey Chia:It's true. Now tell us also about that long form content. I think there, there is a lot of, you know, chatter about long form, um, and people also being worried about using AI to write long form, especially for B two in the B2B space, probably need a lot of blogs and articles out there. What are your thoughts on using AI for that? Um, and, you know, what should people look out for?
Kuba Czubajewski:Hmm. Um, the, uh, the main issue about long form content and AI is that, um, if you, um, you have to work in parts, you have to break it down into, uh, into steps. So the, if you ask Chad GPT, send, send, even if you send a brief for the article, a full beautiful brief for the article and ask it to write it. Right away, it'll not meet the word count. It'll be very weak. It'll be, uh, it'll have, you know, missing points. It'll omit things because it, you know, it works with, with within that context window, even the fifth version of chat, GPT does the same, uh, or Claude or whatever. Uh, the, the way that we approach writing long form content is we break it down into parts. We work section by section. It's still, it uses the same assistant, uh, for the, for the whole article. But we, uh, instead of asking it to write all of it at once, we, we, uh, develop the section. Then it goes through the editor. Uh, once the the section is developed, it goes through the editor. The writer is again, uh, improving it. And, uh, once we, uh, the human overlooks it, uh, and reads the, the, the actual words that are put in on paper, makes the final human edits, uh, adds examples from the clients, um, you know, data or, or examples from their, from their, um, Google Drive that, that our clients gather for us. Uh, and, uh, at as quotes and things like that. And, uh, then only we move on to the next section. So it's a little bit different, but maybe not really, but maybe not really, because if, if I would write all of this content, uh, by hand, uh, I would probably do the same thing. Maybe the editing would be the, the second step after I wrote the whole art, the whole article. But, um. For me, the right, right now, it's important to work section by section and move to the next one. Only after you, you know, that the, the AI did a good job, um, you know, developing it and, and, um. Turning into something useful,
Audrey Chia:definitely even for me, right? So when I craft landing pages,'cause I'm more of a conversion copywriter. If you tell AI like, Hey, craft me a landing page and use a single prompt, 10 out of 10 times is gonna be a very flat piece of content. So even for me, like what you see, I break it down. Steps. I have seven different steps, like audience insights, competitor research, value proposition, tone of voice structure and style. And only then when AI put it together, you get a really high quality output. And I think the mistake that most people make is they take AI content as is, um, without, you know, using that keen eye to look at it and review it. What are some. Mistakes that you see business owners making with AI content. Cooper, I would love to know from your perspective,
Kuba Czubajewski:um, the, the main thing, uh, and the most dangerous thing business wise is not fact-check, fact-checking the content properly. Uh, there is, um, a lot of trust in what AI does and what AI says. Uh, among people who use it for content, and you always have to be skeptical. You always have to be skeptical. Just because the words look good together doesn't mean they, they mean anything. It's a huge risk if you, uh, decide to publish something, uh, that. Just sounds nice and you know it, it's a good, I dunno, mental model or, or a picture in your head, but then someone calls it out on your LinkedIn or on your whatever, wherever you post and says, okay, but that doesn't really make sense. I don't see it working like that within my industry. I don't see working like that, uh, among my colleagues. Uh. That's the scary part of AI and something that I always tell my clients to avoid. Never publish the first output. Always read it. Always think about it if that's something that actually happens, uh, that you, that you observe. The second part part is probably. Not adjusting the content to your own personal style. It literally takes five minutes to read through the AI post and add your own quirks to it. Break down a, uh, break up a few sentences, uh, instead of. Instead of a rhetorical question plus a punchy answer, we write it into something a little bit different, a little bit more original. Uh, I know it's very, very nice of me to say as a writer, uh, and it may sound easy to me, but, uh, with, if you practice it a little bit to just stop for a second and think, okay, but how I would say that, uh, you know, uh, how, how I can, how I can write it so it fits the way I speak to my clients more, uh. So, so that's that. I think that's the most important part, to not lose your own human voice within the, all the AI content that you can create. And the third one, which is my absolute. Least favorite, I hate it with all my guts, uh, is when chat GPT proposes you, uh, the, a very fancy name for something that is happening to your clients. For example, let's say you, you write about a specific problem. Uh, let's stay within the David with my client, uh, within the data analytics and data data management space. And it calls it the, uh, data Frankenstein. Uh, AI uses this figure of speech very often, and when you say on your LinkedIn, I call this data Frankenstein, you don't. You actually don't. Uh, and that's something that I always. Tell my clients to watch out for those fancy names, for frameworks and things like that should always come from your practice. AI can help you name them, uh, but it should be something original, something that you've developed. Describe your process. Ask AI to, to find a fancy name for it, but not the other way around. Not, not do not approach, uh, you know. Claiming a, a term or, or, you know, some, something, something fancy from the place of, uh, because just AI told me to, to call it that. I will just call it that because then clients might reference things that you said that AI wrote for you, uh, that you never said, and you will not be able to tell what you meant because you will firm. You will forget. You will forget that she said something like that, uh, in, in one LinkedIn post out of a thousand that you publish. Uh, and that will be, that will seem insincere. So keep be honest and Al always approach it with more of a cautious eye for, for, you know, uh, finding names for frameworks that you actually use. Not necessarily just letting AI create a. Beautiful, beautiful name for something that never happens.
Audrey Chia:Yes, absolutely. And I think maybe one thing I would add is just because it sounds nice doesn't mean it makes sense. I've seen this happen way too many times. It sounds fancy when you read it, it's like, oh, that sounds smart. But actually when you read the sentences, there is no substance behind it. Mm-hmm.
Kuba Czubajewski:Yeah.
Audrey Chia:Tragedy is really good at writing, um, fancy essays, that's all I can say about that.
Kuba Czubajewski:True, true. That's also the part of the human, uh, the human part of human plus AI system. You always bring in sub substance. Uh, you, you sh you don't have to be a beautiful writer. You don't have to be a talented writer to, to create great content nowadays because AI will do it for you, but it won't bring. The things that you actually do for clients, it won't bring your own point of view. This is something that you need to do and uh, as. As easy as it seems, uh, while using AI tools, you have to do some heavy lifting before you, before you get any decent results.
Audrey Chia:Absolutely. So maybe for, for example, startup, you know, founders were listening to this podcast or someone trying to start a B2B business and they're like, you know what, um, Cooper, where should I begin? How do I get started? Do I need to be on all platforms? Am I gonna be, you know, searchable by chat GPT? Who do I speak to? What do you recommend them?
Kuba Czubajewski:Um, definitely do not try to post on all the platforms possible. You will burn out, you will get annoyed about posting content, and it you will give up after a few weeks or, um, if not days. Um, so what I recommend to all my clients is to build a very simple demand generation system, inbound demand, uh, generation system, which basically means that you should, uh. You do all the heavy lifting work of the defining your, your audience and data. Um, second, you develop, um, you, you pick one distribution channel, which is social media. What channel do you use to distribute your content to spread your message across the web that you're doing what you're doing? Just one. It can be LinkedIn for B2B usually, usually is LinkedIn, but not always. Sometimes it's YouTube, sometimes it's TikTok. I worked with brands that work best if they work on TikTok. Uh, so, so pick one, one distribution channel. Uh, master it. Uh, and uh, within this channel, once you have the distribution, uh, established, you should, uh, create a content, content mix. So a, the, the, the formats that you, that you publish throughout the week in a way that, uh, offer two of them should offer. Value, uh, something very practical someone can implement right away. It builds trust, it builds, you know, momentum for your brand when you're useful. Um, one of them should be social proof. Uh, so a very specific proof that you, that what you're saying and describing in, in the value post is working. And, um, one more personal, so something, something about your company or yourself, uh, that is more. Uh, that kind of builds the, not only the, the value of your account, but also it brings people closer to you. It brings people, it feels like you, uh, like they, for, for them, when they look at your personal post, they should feel like, okay, I want to work with this guy. He seems like a very nice guy or, or a girl. Um. And, um, one of them should be a direct ask for a micro action. And here I'm not at the, you should not ask your clients to book a call with you right away. Uh, no. Uh, nobody has time for that in B2B or in B2C as well. But, um. What I also recommend to all my clients is to create a very small, very, but very practical lead magnet or, uh, an entry level offer of some sorts that they can, uh, that they can implement right away, that they can grab from your profile or your website quickly. Very easily understand that, uh, this is what they are looking for and this is what your pro, your service, uh, uh, addresses. So, um, so those micro action CTAs are very important to one, at least one of your pros should have this micro CTA, um, included. Uh, because, and when you see this, when you have all this system established, you walk your client through the entire buying process from. From the beginning to an end, from top to bottom, whatever you want to call it or visualize it. So value is top of the funnel, middle of the funnel. Funnel is the proof and bottom of the funnel is the micro CTA, uh, so, uh, and each week it repeats. So even if somebody didn't see the post, the the micros d last week, they will see it this week or next week. Uh, and this builds this repeatable demand generation for, for B2B companies that I work with.
Audrey Chia:Awesome. I love the funnel approach. So if you are not thinking about content strategically, this is a great place to start. Think of your content as a funnel where you really have to nurture your audience. And if you're only gonna be selling your product all the time, there is a high chance nobody's gonna care as much. So you wanna give a lot of value at the start just to open it up. So that people will come into your funnel and better understand what you do at off work with that Cooper, that was such an incredible conversation. Now, where can people find you and who should reach out to you?
Kuba Czubajewski:You can find me, uh, on LinkedIn primarily, uh, um, by typing my name, Kuski. I know my last name is difficult, but yeah, I, I hope you'll, uh, you'll manage. Uh, but also you can check out my YouTube, uh, which is Content Kuba, uh, where I post, uh, about content strategy. I post breakdowns of great small companies, B2B companies that are doing great content, uh, purely organically. So, so yeah, you can, you can find me there.
Audrey Chia:Awesome content. Cooper is such a cute name. I will take it on myself. No, thank you so much, Cooper for joining us and thank you folks for tuning in on the AI market display and to our listeners, if you have enjoyed this conversation, share with a founder or a friend. See you in the next episode. Take care.