Power Up Your Marketing With AI Personas [Podcast & Recap]

Ever put together a seemingly brilliant marketing campaign only to see it totally missed the mark in performance?

Do you feel like you don’t yet have a handle on who your target audience is and what their pain points are? What if there was a way to build powerful marketing personas that leveraged AI to not only help you craft them but also draw on the incredible wealth of data and information out there to build more robust and insightful personas for our businesses and marketing efforts? That’s what we’re covering in the fifth episode of the MarTech Show, hosted by head of Strategic Partnerships at Agorapulse, Mike Allton, and co-host and agency owner, Robin Dimond.

You can listen to the entire podcast below. Or read on for a summary of the conversation.

Mike: When it’s time to build out a marketing strategy for business, one of the first steps is always to think about your target audience. Who are you selling to? What are they struggling with?

And it’s a critical step because this is where we make sure that our product and our marketing are speaking the same language as the people we’re trying to talk to. And yet this is the step all too often some of us gloss over because it’s challenging, and its proper personas require that we’ve taken the time to talk to and learn from our target audience.

Compiling and analyzing information from lots and lots of people and determining where they have overlapping interests and aspects sounds like something AI should be able to help us with, doesn’t it? Well, that’s exactly what Dustin W. Stout is going to talk to us about. A serial entrepreneur, Dustin has been in the digital marketing space for over a decade. Throughout this time, he’s focused on building products and resources to help creators and businesses accomplish more.

And his most recent venture, Magai, is his most successful yet, building on the extreme interest and need for comprehensive AI solutions. Dustin has created Magai, and we can’t wait to dive into it.

The Journey to Starting Magai

Struggles and setbacks

Dustin: So it’s hard to talk about how and why I started Magai without talking about the last four years of my life.

Starting in around 2019, I was in a previous SaaS company that was not doing so well, especially from a partner standpoint. I had two other business partners, and we just weren’t aligned on things anymore. And so there was a lot of turmoil in that company. I decided in late 2020 to go out on my own, to leave that company, and to go start something new. Had two great product ideas. And long story short, let’s fast forward: They both failed miserably.

And so here I am at the end of 2022, two failed products. I’d lost thousands and thousands of dollars on trying to bring these products to market, and they just weren’t working. They weren’t finding the product market fit.

And so I was at the end of my rope. I mean, really, was the lowest point of my life. My wife had just gone through a lot of health issues, and I had three kids that we were trying to raise. And things were just really bad. Nothing I was doing seemed to work.

All I wanted to do was just make cool stuff that improves people’s lives. That’s what a lot of us entrepreneurs want. We just want to make the world a little bit better and carve out our own journey as business owners. And I was failing in every sense of the word, it seemed like.

Discovering a new focus

And then ChatGPT came along. I had been working with AI tools for a long time, things like Jasper and CopyAI, all the AI copywriters out there, but they never quite fit in a lot of the workflows that I needed to work in, and they never really hit the mass market.

But when ChatGPT came around, something changed. I realized right away that the world was never gonna be the same again. So previously, AI tools were based on templates. You had to pick a template and fit your idea into somebody else’s box, and it never quite worked. But with ChatGPT, you could just tell the AI what you want, and it does it. And it was magical. And very quickly as a business owner, I’ve been working with content marketing and creating content for over a decade.

Immediately, I found some things that were like, “Why doesn’t it do this? Or why can’t it do this?” Or “I wish I could search my chats and organize them into folders.” And now, wait, Google’s got its own AI thing over here. “Do I have to have two separate tools so that I have to tennis ball bounce back and forth between them? Oh, wait. Now there’s another one. Oh, wait. Now there’s image generation.”

So very quickly, I realized that the world was changing.

These AI tools were going to unlock potential that we as humans have creatively and productively that we need to learn how to use in a professional and streamlined way. And I immediately saw some things that I could do to build an application in which I can leverage the most powerful AI tools, bring them into one interface, and help business owners and working professionals have all the tools at their disposal in a consistent and beautiful and easy to use interface. And that was the catalyst of Magai.

Mike: Love it. Love it. I think that’s perfect. That was good to hear

Robin: Thank you so much for sharing, Dustin, about how that came in. I think as a CEO too, like, finding a solution, you resonated with us a little bit. Like, for me personally, you resonated with me when you said you failed, got up and I did it again, and found something that worked.

And I think that’s something that’s so powerful because it means your tool is even better. You figured out a solution to help so many business owners and how important that is today, especially today. And you’re from personal growth. So I appreciate you just sharing your personal story with all of us

When crushing it, be careful not to be crushed

Dustin: We live in this every day where everything is “perfect.” Everything works out. We’re all crushing it. Like, we’re on the beach, doing business from our cell phone. Like, it’s always good all the time.

I know that there are a lot of business owners out there who are just being crushed. They’re not out there crushing it. They’re being crushed by the weights of their business, the things that are not going right, things that are failing.

And with all these people only sharing the successes on social media, it can be really, really hard mentally. I mean, for me personally, when I was going through the storm, the desert of my life in business, and I did not want to see or hear another person talking about how they crushed it, how they just landed another five-figure or six-figure deal. I wanted to see somebody out there who is struggling through it and can give me just a little glimmer of hope that I’m not alone in my struggles.

You can get through the darkness

And so I take every opportunity to let people know I went through a dark period.

There are dark periods in our business, in our life, and you can get through it. Just keep putting one foot in front of the other. It’s not all roses and rainbows. Sometimes it’s hard, and sometimes it’s a battle that you have to fight, and you’re not alone.

You are absolutely not alone. Social media may make you think that you’re the only one struggling, but it’s not true. Mental health is real. It’s important, and it’s important to talk about and always reach out to people because there are people who care. So I’ll just say that. Thank you for coming to my TED Talk.

Robin: I love that. And this is so important in just how you’re talking about it. And as owning our own companies, really developing, wondering who do I reach out to? How do I reach out to them? And you actually talked about the importance of developing marketing personas. Can you go a little bit more into detail about that? Because like you just said, you’ve got to figure out who you’re even talking to as a business owner.

How to Develop Marketing Personas

Dustin: The challenging thing with all these AI tools is they’re very vanilla. They are trained on large, large swaths of data. And they have all this knowledge, but it’s all sort of blended together.

As a business owner, there are really two two angles that you need to hit your business at. Number one is who am I serving? What is the target audience? What is their personality? What is their avatar? You know, whatever you want to call it. But then there’s also from your standpoint, who are we, and how are we communicating to our audience? What is our brand, so to speak? And what is the personality of our brand?

And so when it comes to generating using AI tools and generating content with those tools, a lot of times you’ll find that the output is very generic. It’s not necessarily what you would create, how you would say it. And so you tend to have to do a lot of follow-up and a lot of editing afterwards to actually get a usable piece of content

AI personas

And, with a little bit of knowledge and understanding, there’s this concept of AI personas, which in which case you can train. And I say “train” loosely because it’s not officially training it, but you can give the AI enough information. You can feed it data about your brand, your business, and have it create content that does actually sound like you, like it came straight out of your own brain, right from your brand. And in the same way, you can teach these AIs, feed it data about your target audience, and either have it engineer content for that target audience or have it act as your target audience and then be able to interface with the AI as if it was role playing as your target audience and get really interesting data out of it in that scenario.

Mike: So to draw a distinction between Magai and ChatGPT, where ChatGPT has got chats, and you’ll feed ChatGPT information within that chat, but it doesn’t know anything outside of that chat. If I understand Magai correctly, what you would be doing is you would do that same kind of a process perhaps, but then you would perhaps name that chat. This is persona number one, and this is the chat where the AI knows everything there is to know about both my business and this specific target audience. And when we go and we want to talk to a different audience, start a new chat, is that right?

Dustin: Not exactly. No. So there is that method. So in ChatGPT, what many AI experts out there will be teaching is how to use a singular conversation and feed it information in that conversation and have the AI learn from that conversation.

Now the trouble in that methodology is twofold.

Problem #1 of AI

Number one: The AI has a very limited memory. Most people aren’t aware of this fact. It can only remember X number of words in a conversation before it has to remove stuff from the top. You know, if you think of an old floppy drive, you know, back in the day, if you’re a photographer, you have a memory stick. You only have so many photos you can fit on that memory stick before you have to start removing some just to fit more in. AI works in the same way. It doesn’t have unlimited memory. Even if you can scroll up in the conversation and you can see the earliest messages, the AI may not be able to.

So at some point, you’re filling it with data and information at the beginning, and then you’re having a conversation. Eventually, the messages pile up and eventually leave the context window, the technical term of that AI’s memory.

So, this is a challenge. If you have vital information that you need to constantly be aware of the persona that it’s taking on or vital information about your business, eventually, that’s going to leave the context window. It’s not going to be remembered by the AI. Now it might be convincing that it remembers because it’s very good at context and might use previous messages to continue that context, but vital information will be lost eventually.

So that’s a big problem and not something you can easily solve with ChatGPT.

Problem #2 of AI

Now the other thing is every AI chat you start in ChatGPT has what is called a system message. System message is simply a set of instructions that the AI has to follow as its most important prompt. So it is, in a sense, a prompt that is given to ChatGPT without your knowledge and without your control.

And that prompt is, in essence, the most generic prompt possible. You are a helpful assistant. And so every prompt you give it thereafter, it will try to follow your prompting, but it still has to view it through the lens of this system message. So everything that you get generated by ChatGPT is colored with, you’re a helpful assistant.

Even if you tell the AI, “I want you to be my expert copywriter. I want you to be my marketing strategist. I want you to be a single mom who’s struggling to make ends meet in rural California.” Even if you give it these personality traits, it’s still viewing it through the lens of a helpful assistant.

Now with Magai, I recognized those two limitations of the memory limit and this system message problem. I want my instructions to be the AI’s filter that it sees everything through. By doing so, that would get me better results. It’s not watered down by some other helpful assistant message. So with Magai, we have this function called personas. And a persona is essentially a system message that is crafted for a specific purpose. And when you engage or when you activate an AI persona inside of Magai, it replaces the system message inside of that AI model. So no longer is it viewing content being generated through the lens of a helpful assistant.

Your persona is now the system message and thereby everything that it generates is so much higher quality. It’s so much more laser focused because you’re not getting this big generalized AI model. You are getting a laser-focused direction for the AI right out of the gate. Now the added benefit of that is the system message never leaves the AI’s memory.

It is always within the context window. It is kept there. So you never have to worry about this vital information, this vital instruction leaving the context window. It’s always there and ever present. Now on top of that, Magai gives you the ability within a workspace. We allow you to silo off projects or clients through this idea of workspaces. Now your workspace can have its own set of custom instructions that is then inserted into the custom persona. So you can activate, say, our expert copywriter persona, but also have some custom instructions set to the workspace. So in my case, I have a Magai workspace. I have custom instructions that detail out what our product is, what our name is, what our website is, the product details, customer, persona, target audience. All that information is vital to my business, now layered on top of the persona.

So you have the most laser focused version of the AI that it’s going to be producing, and it’s preloaded with vital information to your business. And so you really get the absolute best possible AI chatbot or whatever your needs are. And you can engage that same persona and that custom information with all of the world’s leading models. You’re not limited to GPT. You have access to Google Gemini Pro. You have access to Claude and all its models. So you can leverage whichever top AI model you think is best for your needs, and then just layer it and laser focus it with personas and custom instructions.

What are the Best Practices If You’re Doing This to Set Up Persona Development?

Dustin: So when it comes to personas, if you’re using it in the sense of a system message so it’s different if you’re prompting it, there are different tactics and techniques if you’re just using it as a prompt in ChatGPT. Whereas if you’re using it as a custom persona that serves as a system message, you want to change your language a little bit.

Be very specific in prompts

So you want to be very direct in how you craft this prompt. You want to speak directly to the AI as if it were a person. So you want to say, “You are a brand specialist. You help the user in refining their brand value propositions or whatever that task is. You want to be very, very specific. Who the AI is, speak to it directly. You are this. You do this. You do X, Y, and Z, and you help the user with XYZ.” And you can even take it a step further and say, “You do it in this way.”

So for example, if you wanted to create a blog post outline persona who is excellent at creating blog post outlines, you would not only say, “You’re a blog post outline generator. You help the user produce blog posts based on the topics that they give you.” You structure it in this way, and then you can actually give it examples of the structure that you want, and it will follow that structure. It’s actually really brilliant when you use the more intelligent AIs like GPT four or Gemini Pro. These AI models are really good at following structure.

Give AI a template to work from

So you can give them a template, so to speak, to work from, and they will produce content output that follows the templates that you give them. I’ll give you a great example. So one of the failed products that I tried to create was called Social Remix. And the idea with Social Remix is you feed the app a URL, it reads the information from that page, and it creates social media content based on it. I spent probably thirty thousand dollars having a development team build that application, which eventually failed and no one bought it.

But a few weeks ago, I realized, “Oh, I can just make an AI persona that does the exact same thing,” because Magai has a built-in URL reader. You can paste a website link, read the entire page. So I just created a custom persona and said, “Here’s the structure.” You create five tweets, you create five LinkedIn posts, you create five Facebook posts, and you give me five Pinterest pin ideas. And I just put the structure in there. And, five minutes, I had the entirety of social remakes built with just a simple set of instructions that the AI follows, and it produces amazing content. And it actually does a better job than the $30K application that nobody bought.

Mike: So that’s what we’re doing in Agorapulse. We have all kinds of fun things coming soon with AI.

But, Dustin, I wanna get back to Magai specifically because we’re talking to a lot of social managers and community managers.

How Can Social Media Managers Best Use Magai to Help Them in Their Jobs?

Dustin: Well, I think first and foremost, knowing from being an agency owner for the last twelve-ish years, working with a lot of clients, if you are in the realm of client work doing producing content for multiple different brands, it can be really hard to maintain a consistent brand voice across all those different brands that you manage

Save time and do more efficient brainstorming

One great, practical way that AI can help you with that is that if you create a persona for the brand that you’re generating content for, it can be a huge time and creativity saver because I’m a firm believer that, just like time, our creativity is very limited on a day-to-day basis. We only have so much we can give in a day before we’re completely creatively burnt out, and I’ve experienced that a number of times. It’s not fun. So, when you have an AI model that knows your brand voice and knows your target audience, you can have it brainstorm ideas for you, write content for you faster, easier, and with less effort than you yourself could have done. And although it’s probably not nearly as good as you are at writing, having it at least come up with some ideas so that you don’t have to, you can just refine the ideas and maybe make a few tweaks here and there. So much more time and energy saved than having to just start with a blank white screen. For me, the most intimidating thing is a blank white screen. So, social media managers, it’s a lifesaver too, your time and creativity.

Use AI for analyzing content and its results

But in addition to that, you know, just analyzing. So there’s a whole other side of AI that is not just for producing content, but analyzing content and results. One of the things I love about Agorapulse are your reports, your analytics and reports that are so helpful and helping me to understand the efforts that we’re making with our social content. With AI, it’s so good at analyzing lots of data. So, for example, I have pulled Agorapulse reports, and you make these beautiful PDF reports. And because of Magai’s ability to read PDFs to the AI, I can just upload that PDF into the AI’s chat and say, “Hey, analyze all this information for me and tell me some insights from this report.” And it will give you brilliant insights. So you won’t have to read through it and try to figure out how all these things relate to each other. Just have the AI analyze it and then interact with it in a conversational manner just like you would with a really smart analyst and have it give you those insights, without having to do a whole lot of the actual reading and analyzing and and going back and forth. You have a partner, a super intelligent partner who can help you work through that analysis with you.

Mike: I love that. So for those of you listening, if you’re using Agorapulse, you can export your report each month as a PDF. You can have that automatically set up to email you so you don’t have to go run it. You can email you the PDF, and then you import that into Magai, and you can get all the insights that you’re gonna then turn and share it with either your clients or your boss depending on your work structure and say, yeah. Here’s the report. Here’s what we think about it. Here’s what happens.

Dustin: Have AI produce the report for you. Then write the email and say, “Here’s the insights that we learn.”

Robin: Think smarter, not harder. I mean, you guys are setting us everyone up for success, and I think that’s something that is just so much appreciated, especially today. And I love what you said about having a white piece of paper just standing there. They’re just like, “Be creative now.” And I’m like, “Oh, okay. Cool.” I always joked around about that with my parents. I was a creative, and they’d demand, “Be creative now.” Come up with that award winning viral post now, and you’re like, can I get coffee? Like, I I was just doing numbers. My brain doesn’t switch, and then that’s your window. That’s your creative window over. And so to have something that gets you thinking, into that spirit, and inspires you, Dustin. I mean, you’re just dropping real life truth bombs here and business ones. It’s so important just to take some of that word off your shoulders.

Save energy

Dustin: I mean, I was in the trenches. I was there managing. I think at one point, there are some people who are phenomenal at managing multiple different projects and clients. I’m not. That’s why I got out of consulting and got into digital products because I suck at managing multiple things. I think the most I was handling at once was twelve clients. I was producing daily social media content for all of them. This was everything from resorts to banks, to lawyers, plumbers, plastic surgeons, women’s rejuvenation clinics.

It’s all those different things. It’s just a taxing situation. And so when you have an AI model that can just help you spark that creativity and get the ball rolling, it’s amazing the amount of energy that it saves you to pour into the quality.

Here’s the thing I think people mix up when it comes to having AI generate content for you and or with you. It’s not that if you do it right, the end result is higher quality content, not lower quality content. If you are sitting there as an expert and you are good at what you do, then I guess that’s the premise, assuming you are good at creating social media content. When it produces something, you’re just not going to copy and paste it. You’re going to read it. You’re gonna use your critical thinking and analyze what it produced, and you’re going to make it better. And so that means you didn’t have to waste your creative energy on just coming up with the ideas. Now you can take what it came up with, refine it, and you have more time and more energy to make it better. And that’s how AI is, I think, not just a replacement. You people think AI is gonna replace you. That’s not true. Somebody who is good with AI is gonna replace you.

Robin: Ah, Dustin! I say it all the time, probably on record, but AI is not going to replace you. And if a business owner right now is watching this and they want to use Agorapulse and they wanna use Magai, I’ll say this. It is a vehicle to get your employees to work. And so instead of them walking to work, they’re exhausted already. We physically, mentally kick to get to work. They actually took a vehicle that helped them to get to that point. Now you get someone who’s more rejuvenated, who didn’t walk in the rain or in the California sun or in the Miami heat. They got there, and they were happy, and they can go on from there. So using AI, people who use it are going to get to that result a little bit better. It’s not replacing you unless you’re letting it because you’re not using your brain. It still needs critical thinking, and I love that you said that.

Dustin: Yeah. And it’s the same way. Mike, you and I have been around long enough, but we’ve heard the arguments for scheduling content. That was a big thing. People saying, “Oh, it’s inauthentic to schedule content.”

No. You’re missing it. By scheduling out my content, I have more time to be in the community and to respond and to engage. And in that same way, by shortcutting a lot of the hardest and slowest parts of the content creation process, you have more time to feed into making it better.

Mike: That’s a great comparison because also when you’re scheduling the content, you’re leveraging data to make sure that you’re putting it out at the best possible time, ideally. Yeah. And so, yeah, the use of AI to generate really great ideas and copy and strategy and so on, if done correctly, is just a smart use of data as well.

Dustin: Mhmm. Agreed.

Mike: Fantastic. Robin, do you have any questions for Dustin?

Robin: This was great, Dustin. Thank you. I think you answered everyone’s questions if they don’t know how to get to know where you are, which is what Mike is about to say is, how do we find you? How do we know who you are?

Dustin: So follow us on Magai.com. We have a very aggressive road map. We just took on some private investment. We had a number of investors who have been interested in what we’re doing and wanted to add fuel to our fire. Thus far, it’s been mostly me doing everything around here, and that has really kind of locked me up to just building products, and I haven’t been able to do marketing.

So we’ve got a little bit of a private investment, and we have the ability to hire some full time developers at forty hours a week. So our road map over the next three months is going to be crazy. We’re thinking of adding all sorts of additional features right now. You can do text generation. You can do image generation. You can do image to image. We have one feature that allows you to do image to video, but we’re going to be adding text to video.

The whole goal of Magai is to bring all of the AI tools, not just text, not just images, but video, audio, every potential modality into one place. And so we’ve got a lot of things cooking in that realm to bring everything into one interface. And the Magai chat interface is going to be sort of the central hub of any AI thing you need.

And one of the big initiatives that will be coming very soon is a prompt library, so a community source prompt library. If you’re not sure what out of prompt engineer or what prompts to start with, we’re going to be building a vibrant community of high quality prompts because there are some prompt libraries out there that get added to ChatGPT as an extension, and they’re not great. We’re going to have the highest quality prompt library available. We’re gonna have a community sourced library of personas, so personas that other people have created and you can then utilize in your organization. Lots of really cool intelligent features just to make AI easier to use and more accessible and, less clunky, than having used fifteen other different tools for all different things.

Mike: Love it. And when you say audio, are you talking about being able to talk to it like the personal AI assistant on my phone and get audio responses back?

Dustin: Absolutely. Yep. That’s one form of audio. Yeah. Talking to it. So right now on Magai, if you use your phone, you can actually hit the microphone button on your keyboard and record your voice, because most mobile devices have the ability to transcribe your voice in real time. But we’ll be adding that in a more sophisticated way to the desktop interface. And then also if you have an audio transcript, maybe you have a podcast that you recorded and you wanna upload that into the chat. Right now, you can’t upload audio files, but soon you will be able to. And then in addition to that, if you want the AI to speak to you because you’re tired of reading, you’ve read social media content all day for the last eight hours, and you just don’t want to read another thing. You wanna have the AI talk to you instead of having to read what it outputs. We’ll be having an AI voice as well so that it can talk to you.

Mike: Very cool. Very cool. Well, that’s all we’ve got for today, friends. Don’t forget to join the after show discussion in the Social Media Pulse community, and that’s where you can learn more from today’s guest and connect with other folks who are attending. Any final thoughts, Robin?

Robin: No. Really appreciate this, Dustin. Thanks so much for joining us today.

Dustin: Absolutely. It was a lot of fun. Thanks, everybody.

Thank you for listening to another episode of the MarTech Show hosted by Robin Dimond and Mike Allton, powered by Agorapulse, the number one rated social media management solution, which you can learn more about at Agorapulse.com. If you want to make sure you’re part of our audience during a live weekly broadcast, take a look at our calendar, or click the subscribe button in your email once you register.