50: How to Teach AI to Speak Your Brand Voice

Your marketing team is probably already using AI. The real question is: are you getting content that truly sounds like your brand?

In this episode of Practical Rebels, Logan and Veronique unpack how to train AI the right way. From giving better prompts to refining outputs in rounds, they share how to turn AI into a creative partner instead of a quick shortcut.

🎙️ The episode is now live. Listen today on your favorite podcast platform.

Transcript

Cold Open & Topic Setup

SPEAKER_00 0:00

Welcome everybody to today's episode of Practical Rebels. Verique here, and I am joined by Logan.

SPEAKER_01 0:08

Hey Varionique.

SPEAKER_00 0:09

How's it going?

SPEAKER_01 0:10

It's going well. I'm up here in the content room in the dark. So it's fun, perfect for Halloween.

SPEAKER_00 0:16

I love it. I'm hiding in the well, not hiding. I'm in the conference room and just closed the doors because it's so pretty outside today because Halloween is coming and it's cooling off. So yeah.

SPEAKER_01 0:26

I know it's the best time of year.

Why Marketing Teams Rush Into AI

SPEAKER_00 0:28

I love it. I love it. We have like two months of wonderful weather and then we're back in the swing of it. All right. Anyway, I am really excited to talk today about our topic. This is something that we have been, I mean, feel I feel like Logan, you've really been hyper plugged in and excited about all the things that are happening with changes in AI over like, I'd say the last year.

SPEAKER_01 0:53

Yeah, I think that AI, obviously, the topic itself is ubiquitous in any community right now, professionally or socially. But I think specifically with AI, I think it just dovetails into so much of what we're doing right now that it's sort of it's just become part of what I keep up with just by default. You know what I mean? Like because so many things that I do here in for clients are impacted by AI.

SPEAKER_00 1:21

So yeah, exactly. And I think we've been really thinking through how to use it uh wisely without watering down the work that we do, without watering down our own brand or our clients' brands. So it's something that we've really been, we're we're always talking about. I'm actually gonna be doing a talk at OntreCon, if anybody's interested in hearing more about this, on November 13th, bright and early. I'll be the first speaker up at 8:45. So definitely grab your tickets if this is something that you are interested in hearing more about. So, what I'll be talking about there and what we're gonna touch on today is really how organizations can maneuver the world of AI to use it alongside their brand for their marketing efforts without, you know, just becoming overwhelmed, becoming burned out, and using it wisely as I like to call it like the brand co-pilot, as opposed to letting it take over and completely kind of dilute what you're putting out in the market. So let's start from the top, Logan. A lot of marketing teams are experimenting with AI right now. They're figuring out what works for them, what doesn't. Why do you think so many are jumping in?

SPEAKER_01 2:35

Yeah, it's a good question to start off because of the why. And I think that it, like I said earlier, whenever we first started, I think that marketing and you know, specifically marketing rather in social media, AI just dovetails so nicely into so much of what we're doing. A lot of the platforms have an AI component. You know, if you're in social media marketing, you know, all the scheduling platforms, they're using AI. You know, Google is completely being, you know, like it's halfway AI at this point. So I think that it's just in what we're doing in marketing, just day to day, it's so AI has been become so ingrained in such a short amount of time that I think that a lot of teams just feel pressure, right? And maybe that pressure not even in a pejorative sense, but you feel like if I'm not using it or if I'm not at least trying to use it, am I in danger of you know, falling behind or missing out? So I think a lot of those teams out there, they're just they're looking for an in and they're looking for an edge to try to keep up because that's what marketing is, anyways, right? Like you want to try to to beat the competition. So this is just part of it.

SPEAKER_00 3:48

Yeah. And on the other side, we're kind of seeing we've used a few different social media scheduling platforms over the years, and we actually got a notification from one of the prior ones that we had used for probably about 12, 18 months that they were shuttering because they couldn't keep up with the integrations of AI with the large ones with Sprout and things like that. So I think the pressure is definitely coming from both sides.

SPEAKER_01 4:14

Yeah, it's definitely coming from both sides. And I think that the platforms that are able to keep up obviously are the ones that are more, you know, established. You mentioned Sprout. So I think that they're able to keep up at a pace that maybe some of these like smaller, more boutique brands um really can't match.

The Brand Voice Problem With AI

SPEAKER_00 4:32

Yeah, yeah, exactly. So when we talk about AI specifically and brand, I think that's one of the places where it can get a little bit tricky. I think that you know, we definitely don't want to be using robots to completely take over our brand voice or lead us lead us astray off of what we have established as like what we are, the human story behind it, what we're about. When we are seeing brands that are when we talk about AI and like the brand voice, like what is usually going wrong? Where are companies going wrong when they are using AI to amplify their brand?

SPEAKER_01 5:08

I think where I see most of the problems with brand voice in AI, and what I think it stems from is, well, first of all, the the symptom is that a lot of the content out there is starting to sound the same, right? You know what I mean? And then it's it's very like it feels or it reads like AI. And I think that that's not necessarily the problem, isn't it with AI, and this isn't me pointing fingers at any brand managers or marketers, but I think that usually the problem is in the chair of the person using, you know, the AI platform. So I think that really it comes from maybe a lack of understanding of what all the capabilities are using some of these LLMs and how to feed it the information the right way so that you don't end up sounding like 10 other or your 10 other competitors out there who are all doing the same thing. And we know, I mean, just working in brand, you know how devastating that can be because if you're not setting yourself apart, then what is what's the value proposition, right? Like what makes you different from any other brand really at that point.

Build A North Star For Consistency

SPEAKER_00 6:18

Yeah. I want to add on to this, but just something we noticed yesterday that when you talk about everything sounding the same, we'll every now and then we'll use like upwork for a small task if we just need a quick freelancer for something. And we'd posted a job for something, and I got within 10 minutes, I had six responses back, and they were all clearly using AI because it was the same cadence of repeating back like the request in a very specific way. It was creepy. Like I dropped it in Slack, and everyone, I would, you were like, I hate this, I want to go back or whatever. Cause it's like lost all sort of like personalization or human touch. And it was just kind of one of those funny things I noticed. And that was just a small example of what we're seeing amplified across like content as a whole. Just this risk of sounding the exact same as everybody else. So what we always say is because we do use it to kind of like build off of base content that we create, that Logan creates, that Human creates creates, and we use it as a way to kind of start to build out different, different like segments, like segmenting, like email versus maybe helping support a social post or whatever. But we always start by like heavily understanding what the brand voice is, what the organization is about. Like it's the brand guidelines, right? And it's what we call the brand guidelines. It's like the North Star document that has like, what is our purpose? Who are our audiences? What do they need from us? And then actually giving it like intensive amounts of content that's actually been created that is the best in class. So it thoroughly understands like everything about what this organization is, how it serves its customers or its clients, and how we communicate with those people. Like, what are their pain points? What are we trying to solve here? And then when you are going in to actually like build core content, you're building the core content, you know? And then you can have it like scan for, hey, is there a better way to say this? Or is there something I'm missing in this? Or go do research for me on this. But it is like you have to like truly know what you're about, what your organization is about, what the purpose is, what your audience needs, and how you solve those needs before you can even like open up a chat window.

SPEAKER_01 8:39

Yeah, I think that's that's spot on, Verenik. And I think that honestly, you know, not to toot our own horn, I think that we understood that at a level that maybe other teams didn't just because, you know, Hatchmark was started as a brand agency. So we've always taken a brand approach to everything. And I think that that really helped us whenever this, you know, the proliferation of AI started, because we understood from the from the get-go that, you know, we have to have, you know, a foundation. We have to know how to talk about ourselves. And in a real sense, if you're treating AI as you like to say as the co-pilot, this is another member of your team essentially who needs to be brought up to speed. You know what I mean? Like that person, you know, in AI in this case, needs to have the brand guidelines. They need to have the North Star document, they need to know everything that you know so that whenever you put in whatever your inputs are, you're getting good stuff out that's actually usable. And you can't do that without doing the work up front. Like it takes a lot of work up front and a lot of original content just to get it to where you want it to be, so that it is keeping that brand voice that's so important.

Custom GPTs As Shared Team Memory

SPEAKER_00 9:52

Yeah. And one of the things that we've started doing as well is building out and something that we had this kind of realization about is we have all these teams, marketing teams, brand teams across all these organizations, and they are all using ChatGPT or Cloud or whatever they use, right, for their own purposes to help them do their jobs. But who knows how trained they have, how how much training they have given them, right? So utilizing something and setting up a very robust custom GPT for your company and your brand, you can then share that with everybody on your team who was creating content. So like the heavy lifting is done up front. And then, of course, teaching people like the ways to and to not use it. But that really helps to start building consistency and also have one singular brand voice that like whoever is leading your voice, your messaging, your positioning has set. And the other thing too is that like with any brand, how you speak and what you talk about is going to adjust like naturally over time. Your audiences might shift, your products and services might change. You have to continuously train and teach whatever AI you're using to evolve as you do. It's definitely not like a set it and forget it. I mean, I think a lot of times we're so used to in marketing, like we know these changes that happen. Like we have to then go in and say, this is a new product that we're offering, blah, blah, blah. Like add it to the list of resources that it has so it can be brought into that like full picture as content is being developed.

Maintenance: Keep Training Your Models

SPEAKER_01 11:27

Yeah. And that's something I've run into now using AI. I would say that AI has been part of my daily process for, I don't know, probably eight months now. I mean, I started using it probably a year ago, but I've really threaded it into my daily process more recently in less than a year. But even as you mentioned, things change, you know what I mean? Like best in class changes, you know, maybe month to month, especially with a new client or maybe with a new brand. So you do have to go back. And I caught my went back in some project files for, you know, one project I was working on and realized that, hey, this, these content, you know, examples that I've used are not necessarily the best anymore. So I need to clean this out and re-upload new content and maybe like give more clearer instructions moving forward so that it's the best practices are the most current. And it's like what it's like anything, right? You have to give maintenance to your website, you have to do maintenance on, you know, you constantly have to be, you know, checking for, you know, ways to improve in social. You guys have to always be checking for ways to improve with your AI.

SPEAKER_00 12:34

Yeah. It's funny because when you start to think about it as like your teammate, or I think Brian, like Vista Techworks Brian, like he was like, Oh yeah, it's like my like cheap intern that like is really good at all these things and can remember and recall everything. We train our teams and keep our teams on the loop that are in things that are going on with the business. Like you have to do that with the technology as well. So just kind of thinking about it that way.

SPEAKER_01 13:02

No, it's weird, and I I try not to get too dystopian with it, but I actually it's in it not in a creepy way, but I talk to it kind of like a human being because I felt like that helped me understand that this is, you know, it's not leading me. It's just, it's genuinely almost like having a conversation with another creative in the room. And I think that's the best way to use it because it feels more natural, right? It doesn't feel so rigid, it doesn't feel like you do this. It's just have that free-flowing conversation. I feel like that's whenever I've gotten the best outputs for sure.

SPEAKER_00 13:35

Yeah, absolutely. How often are you when we talk about keeping it up to date, feeding it examples, content, like best in class? Are you like regularly loading new articles that you believe have been really successful, or is that all within your project group and you're just kind of referring between? Because the amount of like recall and cross-referencing it can do now, I have found is crazy.

Treat AI Like A Teammate

SPEAKER_01 14:02

Yeah, I've noticed, you know, I think early on, whenever I was first using it, the recallability and the archival knowledge is very impressive and it gets better as time goes on. But I have realized that if I start, like my baseline is more or less going to be the same for any one project. However, if I'm going to start something new and I know that there's new stuff to talk about, I might use supplemental content in that one to-do or that one task to make sure that I'm re-explaining things. Because sometimes I'm just like, if I want to say, hey, do you remember that thing from three months ago? Like the chat might come back and say yes, but it might be a little bit muddy. So I'm I would say almost every single time I sit down, I'm giving pretty detailed prompts and maybe even adding new examples just on top of what's already there, just to make sure that I'm not wasting time or I'm not gonna spend a ton of time trying to edit down by not giving myself the best possible chance for success on the first go or the second go.

SPEAKER_00 15:12

That's really funny because, like what you just said, if you're like referring to something from three months back, if you were having that conversation with a human, they would not remember exactly what you were talking about. It's basically like you just said, you have to start treating it like a coworker or this, you know, supplemental person in the room to help you out and like remind it.

unknown 15:38

Yeah.

SPEAKER_00 15:39

Yeah, it's weird.

SPEAKER_01 15:41

Yeah, no, I even I'll even say things like, Do you remember this? Yeah. I guess it's whatever, but I just I'm always trying to like give more supplemental context. And I think we'll probably get to this at one point in the episode where what you put in needs to sound and feel a lot more like a true creative brief and not just a hey, write a social post for me because I it just it's not gonna work like that.

From Prompts To Creative Briefs

SPEAKER_00 16:09

Yeah. So let's talk about that. So you're alluding to, I believe, like prompting, right? Um, okay. So what makes for a good prompt versus a weak one?

SPEAKER_01 16:20

So what I've and I I I already mentioned this once, but the way I like to treat my prompts is as like creative briefs, and anyone out there listening who's in marketing or advertising or whatever is familiar with a creative brief, but really it's in layman's just you want to give it as much context of the project. You want to say, like, who's the target audience? You know, what's the source material? Why are we doing this in the first place? You know, what is the intended outcome of this? I mean, sometimes I'm sitting there and I'm typing my first prompt for maybe 10, 15 minutes where I'm just really trying to give it as much context as I possibly can because it's kind of like garbage in, garbage out, right? If you put in one sentence and it's lazily done and you don't give a ton of context, it's gonna come up with something for sure. Like it's going to give you something, but it's not gonna be great. And it's probably not going to be contextually correct in what you even need. And you're gonna spend more time trying to fix it or like zhuzh it up here and there. So I feel like the more time you spend on the front end, you can save a lot of time and just spend your back end time doing small edits because even if you do it all the right way on the first go, you're still gonna have to edit, right? Like you're you're it's not going to be a copy and paste game as much as people might want it to be right now. It's just not there yet. Maybe it will be one, but I've haven't seen that. And um, yeah, you're gonna have to like you're gonna have to spend a little bit of time up front and make it a true creative brief for the prompt.

SPEAKER_00 17:59

Yeah, but I think like I I don't think it'll get there. I think you're always gonna have to feed it that high quality information because again, if you were to if I was gonna walk over to to Joe and just say, hey Joe, here's one more like sentence of what I want you to do, go like without that additional context, a person's gonna not gonna know how is your AI gonna know.

SPEAKER_01 18:22

Yeah. It's almost like having like a miniature meeting with it beforehand, you know.

SPEAKER_00 18:27

Yeah. And one of the things that I've noticed is really helpful too is I can you know put in the initial information and then say, what other things do you need to know from me to best give you, like to give you the best input, yeah, and almost have it prompt me to prompt it, if that makes any sense. Yeah.

SPEAKER_01 18:48

Yeah, no, that's a great one. Because even if you don't do that, I've noticed sometimes it will ask you for additional context without being prompted for context. You know, it will say, like, hey, I feel like I understand it, but I I need a little bit more context.

Back‑and‑Forth, Practice, And Standards

SPEAKER_00 19:02

Yeah, 100%. How much back and forth have you experienced? And has this kind of back and forth changed as you've gotten better at utilizing AI?

SPEAKER_01 19:15

Yeah, it's um well, for one, yes. I I think that I think AI is such a new thing, and we treat it as this like this thing that no one's ever seen before, which in so many ways that's true. But I think that there's a lot of just like fundamentals of anything that apply with AI, which is you practice makes perfect, and maybe you're not gonna get perfect, but you can keep using it. And I only say that to say that the more I've used it, the better I've gotten at it, and the less back and forth I've had to do with it because I'm getting what I want earlier without having to do a ton of back and forth. But as I alluded to earlier, I don't really think, as far as I can tell, that anything has ever come out the exact way that I want it to on the very first try. Maybe, and some people might get it, you know, and maybe I just, you know, we set a high standard for ourselves. I think most creatives do. So it's just like it almost feels wrong to just be like, yeah, first try, that's perfect. I mean, there's always ways that you can read give it a close reading and say, you know what, I think this could be better. I don't know if this quite aligns with our value prop, whatever you're doing, really. So I think back and forth should be expected. It should be encouraged, but you definitely can trim down how much time you're spending on that by practicing and getting better.

AI As Force Multiplier, Not Idea Engine

SPEAKER_00 20:42

Yeah. Yeah, exactly. One of the things I want to talk about a little bit too here is this idea of I think a lot of people, at least at the beginning, and I think it's changing a bit because I think people are realizing that if they don't put the time into it, they're getting, you know, stuff out that is homogenized. It sounds like everything else, it's not great. But the approach of looking at AI as a shortcut and getting things done faster, I think is fading. What has been your approach to look at it as like a creative partner to help iterate to help? Like for me, it helps to free up my brain so I can spend more time on the upfront, like big idea, the content with purpose and intent, and like providing a piece of content that's valuable to whoever's opening, you know, in our your case, like our email or a social post or whatever it is. Have you had that same experience where you're you just find yourself spending more time on the purposeful content and then being freed up and being able to utilize it as like a distribution system?

SPEAKER_01 21:51

Yeah, absolutely. I think that it's it's interesting because I've been thinking about this recently, where it 100% is a shortcut in any capacity that you're using it, but it could be a shortcut to something terrible. You know what I mean? Like based on what your what your intentions are, it could just be a shortcut to something that sucks. So I mean, I think that you have to be using it the right way. And as you were speaking about just having that core content, what I find is that I'm still spending roughly the same amount of time developing the core piece of content that I really want because you have to come up with your own ideas, and I think that that's so important still. Like if you're trying to use AI to come up with everything for you, you're going to be in the same boat as everyone else, and you're going to sound the same. But if you're still coming up with your own ideas and you're still coming up with your own content ideas, using that and then having that core piece be the thing that gets you to the next four or five pieces and using AI to help you shortcut to that endpoint, I've had tremendous success with that. But it's just I can't, you have to underline it, highlight it, encircle it. That you you have to come up with your own original ideas and you have to come up with your own thoughts. And because otherwise, like, why would anyone want to read that? Or why would anyone want to even look at that or consume that content if it's just if it's junk? So if you're a a thought leader in any capacity whatsoever, don't use AI to replace what ideas would otherwise be great coming from your own brain and try to replace that.

Research Faster, Think Deeper

SPEAKER_00 23:39

Yeah, one of the ways that I've found that I enjoy it the most is kind of that tedious stuff that you know, unless you have days to do one task, you can't do it super all. So, like, you know, when we using deep research to go in and do competitive research or pull data, guide it on sources, have it called articles and pull out the big ideas and then do a skim of those articles, like double check on it. But the time-saving stuff there so that research portion can be supported has been like my favorite thing lately. And then I'm able to take that and be like, okay, then what does this mean? Like, what does this mean for our client or for our brand or whatever it is?

SPEAKER_01 24:19

Yeah, researching citable sources has been great. I mean, you used to have to just like call the internet um and hope you're on the right, yeah, the right place. Just like in hope that you're getting something that's usable, but that's been a huge thorn in, I think, everyone's side for the longest time is trying to find something interesting to cite or something like that. If it's a written piece of content, AI's been, you know, magnificent in that respect for sure.

Leading Teams Through Adoption

SPEAKER_00 24:47

Yeah. How if there are marketing leaders or you know, groups that are considering starting to incorporate this more into their daily workflow, how would you encourage them to like get their staff more comfortable with using these tools and not feel overwhelmed and not feel like, you know, they can't figure it out because there is so much going on with it constantly?

SPEAKER_01 25:15

Yeah, I think it does. I think leadership and encouragement, obviously it starts from the top. So if you want your team to use AI and you're not willing to use it yourself, I would question that. You need to be leading by example. You need to be saying, like, hey, here's what I'm using it for. Try this. And to become an expert at it yourself, you know, like there are people who, you know, for a lot of good reasons, are hesitant to try it. And whether that's a social thing or whether that's fear of job security, I mean, you go down the list, these are all very real fears, and these are all very, you know, good fears, I think, to stop and ask ourselves, you know, are we, you know, are we doing the right thing here? So I think that just having open and honest conversations with your team and just uh encouraging and letting people try new things. And if you are someone who's still on the fence or you're just not comfortable with it yourself, you know, you really probably aren't going to foster an environment where people are using AI because they feel like they can't. So I think really just opening up that door and having it's just like, hey, let's not, you know, be ashamed of it or anything like that. Like it's this is a tool that we should be using and just make it a free, you know, a free market kind of experience where people feel like they can use it and express themselves and bring new ideas to the table.

Start With AI Inside Your Current Tools

SPEAKER_00 26:41

Yeah. And I think that also, because as I was, as I've been working through this, this entrecon thing, I realized that when you start to focus on having a good understanding of the tools that you're already using, this is going beyond cloud or chat GPT or anything like that, but like your email tools, your social tools, you know, how Google is scanning for search, utilizing AI, like the tools that you're already using, get familiar with AI integrations and how you can leverage those within that specifically, and then kind of build from there. Because I think a lot of time it's knowledge that builds on itself. And if you start in one place, then it becomes easier and more comfortable to be like, okay, now this has kind of like helped my life, helped make my day, my workflow, whatever, a little bit easier in this way. How can I layer in the next thing? So instead of trying to like go all in at once, like pick a few tools that are already using it that you can start to like lean into those, those ones specifically.

Encouraging Everyday Experimentation

SPEAKER_01 27:46

Yeah, and totally like noodle with it, you know. Like if you're new to it, the way that I became most comfortable with it is using it every day and just trying new things. You know, you're you're not gonna get acquainted with it, you're not gonna get better at it by not using it. So, and like you said, there's so many platforms that have some AI component, even if you're not completely sold on, you know, some of the other ones, try using AI for some of these other things. And we use Clavio in the emails. Segmenting with AI has been great. There's just so many different things that you can use it for, and it's there's there seems to be a new capability launched pretty much daily on these platforms. So just try it out. And there's always a little banner popping up on your screen talking about try the new AI, but so try it, see what it's like. You might be surprised at how how much it helps.

SPEAKER_00 28:36

Yeah, 100%. I think we've covered some really great stuff here. I have been surprised at how seamlessly like it has started to work into our day-to-day just with familiarity of use and just you know making the kind of redundant tasks easier too. Is there anything else you want to add to this?

SPEAKER_01 29:00

No, I think this is good. And I just, you know, like I think we've had a success with it because we've had an environment that has encouraged it. So if you're listening to this and you're on the fence, or if you're a part of a bigger team or just trying to a small business owner trying to get some more out of out of what you're doing, just try it. I mean, you really don't have much to lose if you just give it a shot and see what happens. And I really think a lot of people will be surprised how much it can really help without sacrificing, you know, that that brand voice that's so important.

Closing Thoughts And Event Invite

SPEAKER_00 29:34

Yeah, exactly. Great buddy, close it out, Logan. All right, if anybody wants to hear any more of our thoughts around how to wisely utilize AI and implement it into your grading and marketing strategy, we'll be at Punchon. Um, it is from November 12th through 13th, downtown Pensacola. So you can grab your tickets at PunchConPensacola.com and looking forward to seeing you there. And yeah, we're always down to talk about this stuff. So if anyone Has any specific questions or has found any ways that they are implementing it in a way that's unusual and different and cool, like drop us a note. We'd love to hear it. Anyway, till next time, thank you so much and see ya in a couple weeks.

SPEAKER_01 30:15

Bye.

SPEAKER_00 30:16

Bye.