The Creative Sidekick: How AI Agents Are Changing The Game For Marketers

Cheryl Boehm: Hello and welcome to another episode of the Leader Generation podcast, brought to you by Mod Op. I’m Cheryl Boehm, and today, I’m here with Tessa Burg. We’ve been talking a lot about AI here at Mod Op, and as we close out 2024, we’re getting ready to launch our Mod agents suite and it’s designed to accelerate AI adoption for both our team here at Mod Op as well as our clients. But not all of us have your technology background and expertise, Tessa. So before we really get into that and into the details, I’m hoping that you can just tell us a little bit more about what agents are and the role that you see them playing in marketing.

Tessa Burg: Thank you, Cheryl. I am excited to be a guest. I actually really enjoy sitting on this side of the table, if you will. So agents are going to be one of the technologies that most dramatically change the role of marketers. I think a lot of people right now are still in that space of testing apps and feeling inspired by them, but as we’ve talked about on past episodes, there is a lot of companies still not able to show the value. About 74% of companies still have not been able to measure and show real value and impact of AI. And agents are a great vehicle for not just inspiring, but going above and beyond what apps can do, partnering them with apps in order to create innovation that is specific to your company and truly specific to where you see the people on your staff and your team members evolving and growing. So what is an agent? It’s an automation is like the simplest term. I think one of the questions I hear a lot is, you know, what is the difference between GPTs and agents? And if you’re not a developer, you can create a GPT. And in fact, I think a lot of marketers, if they haven’t created a GPT yet, give that a try. Those can be hyper specific to things that you wanna do. And the way that you would create one is exactly the approach that we recommend for clients today. Think about what you’ve always wanted to do. What are some areas that you’re spending a lot of time doing research or compiling data and that research that data, your past presentations all become the brain or the fuel that feeds the brain of a GPT. And then in that regard, it operates very similar to ChatGPT. So if you understand what ChatGPT is, then creating your own GPT isn’t too hard to undertake and really not too hard to understand. And then agents may use GPTs and they be, the GPTs can be the brains for the agents, and sometimes, it’s a combination of GPTs that power an agent. But the biggest difference is AI agents are autonomous and they can perform specific tasks and achieve specific goals. So if we think about how we’re creating GPTs, we’re thinking about, we’ve always wanted to do, what kind of data do we have to power that? And an agent, we might be laying out a process, so now because I did this action or because I got this output, what do I want to do next? And that makes them a little bit more broad in scope than GPTs and also gives you almost like I always hear this phrase now that, you know, everybody wants more passive income. That’s what agents do for marketers. They become the passive income, they’re taking the actions when marketers are now being elevated to spend more time on solving problems and strategic thinking.

Cheryl Boehm: Thanks for that explanation. That was very easy to follow, especially for us, you know, that don’t understand all the technology all the time. It’s interesting, I heard this adage once that you should never recommend something for a client that you’re not willing to try and do yourself. And that’s always hit home with me. I always thought that was really important. And it’s important to note here also that Mod Op is beginning to use agents today. So how are we using them as an agency and where are we seeing the most value from them?

Tessa Burg: Yeah, so we started our AI journey like way back in the day. And Cheryl, you know this because we were together. And it was about 2016 is when we gave our first presentation on how our clients could start to use their data to power hyper-specific automated workflows. And those workflows were to give insight and findings that the clients had not had access to before. And you’re like, but why? We had Google Analytics, we get performance metrics from how well our marketing is doing, but without sort of diminishing those metrics, they really are like vanity metrics, like how much traffic you get to a website, how many impressions you get is not always indicative of results. And our clients, like a lot of clients are always pushing for, you know, show me the money, what’s the actual ROI? And so back in 2016, we started looking at, okay, what data gets you closer to measuring the true impact of generating leads, of running ads, of doing an email campaign? And what we found is, sometimes, it sits somewhere else, outside that marketing sphere. And with the advent of machine learning, we were able to more easily bring that data in and create models to better understand what segments of customers and what their behaviors were by segment, and start to manually personalize campaigns. But all of this, all of that was manual. One of our first presentations was looking at how do you take IOT data, pull it in, organize it, what do we know about it? How does that start to show when something was purchased and when it might need maintenance? I mean that was like a year long project. And then about two years after that, we found a platform that automated doing attribution and allowed us to bring in that external data. But again, setting up those campaigns, super manual, like you had to know who the segment was, we had to put in all the creative, and very, very heavy lift. So from those experiences, we understand the amount of upfront effort there is to power an AI solution. And now with the democratization of OpenAI’s API and other LLMs, we can create this wishlist. Man, I really wish we would’ve been able to get to that result faster. Man, I really wish I would’ve had that data and those insights before I started this campaign concept. And so we’ve been able to use what we have built since 2016 to be really data-driven, but it’s been a journey. But because of that journey, we now have very specific opportunities to automate and pull that data upfront. So what we have found when we’re thinking about our agents heading into 2025, it’s how do we start to get data about the customer, about the brand and about the landscape upfront, in front of our concepts, so that we are getting these predictive insights that inform opportunities, that inform the customer journey, and give us a faster way to show the client new ideas. Wherefore it was, you know, just, I’m sure everyone has run into this just like pulling data, populating the Excel sheets, and now our agents are really becoming more of that creative sidekick and giving us insights and visibility that we simply haven’t had before. But it starts with like looking at your use cases of where are you spending the grind and what would it look like if those grinding activities were automated or if they moved entirely from the bottom of the process to the top.

Cheryl Boehm: You mentioned a lot of these experiences that we’ve started with since 2016. Are those experiences largely what inspired the development of this agent suite or are there other factors that played into it that helped inspiring our creation of these agents for both the creative team and all types of marketers?

Tessa Burg: Yeah, that is a great question. So when ChatGPT launched, we immediately launched the AI Council and we’ve talked a lot about that and preview, we’re gonna write a book on how to stand up an AI Council, because that was a whole other thing in itself, and we learned a lot. And one thing that came out of that was the inspiration for the agent suite. And it’s because we brought such a diverse group of people together. So the example I gave that’s, you know, that’s our lens, that’s our experience. And when we brought more creative folks in, more video production, more media people from the analytics team, we started to realize that there were some intersections into each of our pain points in each of our experiences, and that connected tissue and the need to bubble that all up to shared goals and a shared vision is where the agent suite came from, and that I would recommend as a starting point for anyone. I feel like if we would’ve only used our experience and started to create agents, honestly, it would’ve been a bit narrow. I think we would’ve stayed in that campaign optimization line, because you know, we started from an agency that was more B2B in generating leads and it would’ve been awesome. And I’m sure if everyone stayed in their own lanes, they would’ve created some super awesome agents, as well. But what we would not have had is a single platform which, you know, we’re calling the Mod Engine that connects the brains across our agency into a single repository that can power multiple use cases per agent. For example, we’re gonna be launching the brand agent. We recently moved it out pilot, and it was very unexpected how many use cases came out of that, but it’s because it’s tied to a platform with a large rich repository of data. And so a single agent can satisfy many different use cases across the organization. And when that has happened, that inspires more agents. It’s like, oh, I use the brand agent, I’ll give a real life example. When it first started, it was more in line with what I said earlier, inspire creative, each client gets their own brand agents, it uses our data about them and their industry paired with data that they’ve provided us, but someone in the evaluation team was from sales. And so they use the agent to take on the role of someone on the client side, and help them better understand the needs and wants of that client side role and where there were gaps like basically asking it what aren’t we doing it as a company? And I was like, “That is thinking genius.” And especially because you know what, you know, any client, any brand can make agents for themselves. But what makes our agents unique is what makes agencies unique is that we work across many clients, and we’ve been tracking usage of our own tools and we’ve gathered stats and insight about how certain types of campaigns work across and in that industry. So not, you know, the data we have is not client specific, but it is very specific to industries and the expertise that we, as marketers, bring. And so I thought the idea of looking at that from the client side and saying, what’s missing, where’s the gap, was amazing, but the question was coming still from that specific client. So that answer can be different for every client. I mean, it’s just extremely exciting, But I would say the, you know, the inspiration, the short answer is, sorry, that rambled there, but it came from the AI Council, and it came from bringing diverse perspectives from the AI Council together. And the continual use of our responsible use policy into our evaluation framework really drives an innovation roadmap for where other agents can then take one automation and make it two to three, and continue to make that connected tissue be enterprise wide.

Cheryl Boehm: That’s great. And as you mentioned, there’s use cases across the entire organization and you gave the example of sales, but as you know, I come from more of a creative background, so I’m gonna be a little selfish here and ask how does a brand agent help the creative team, and help ensure that creative outputs remain on brand and really align to the client strategy?

Tessa Burg: Yes, I think in the creative process, we have so much rich data from what’s worked in the past from who the customers are, we also then supplement it with data that we’re licensing from other sources around what is trending. And since we do have these highly specific agents that automate thought on behalf of the client, we’re able to pretest our ideas. So in the creative process, instead of wondering, “Oh, I hope this concept really does well,” you know, like you can actually get that forecast of impact by customer segment upfront and clients still have feedback. You know, we’ve tested this a couple of times, so clients are still gonna come to you and be like, “I see what the result might be, but I hate blue.” And that’s fine because clients represent their brand, and it’s at the end of the day, their brand on the line. And that inspired another good creative tool, which is brand safety. So we can pretest our creative for performance, we can pretest it for brand safety, making sure that where it’s going to appear in channels is in line with the brand’s values and the audience. Those have probably been two of the biggest use cases that expedite the approval process instead of getting caught in subjective conversational loops. And plus, clients really love to take that data back and communicate internally, and get everyone excited about what’s about to launch. The other one is looking at a single creative concept across a journey and helping us tighten up the different ways that we engage and create feedback loops with the client’s customers and their target audience. And that’s been very inspirational. So here’s this campaign, it’s running on these channels, what are the opportunities we have to get more feedback from our customers on a website, on their mobile phone, in the app? And so I think while we always thought of creative as cross channel, we’re able to get more specific and tie it down to really meaningful KPIs in that pitch process, which then expedites the rest of the process. So now when I hand it off to development, they know exactly why they’re putting that CTA, what it’s tracking, and everything just becomes a little cleaner. Like I feel like if we had this back when I got my agile training and software development, that would’ve been awesome, because you know, as a developer or sitting in the engineering side, marketers, sometimes, would say these very big broad statements and I interpret it one way, and they’re like, “That’s not what I meant.” Where now we’re tying things down to the exact type of data we wanna learn about them. And the why is crystal clear and I’m seeing the entire journey, woo, that’s like development gold. So seemed very simple to be inspiration at the front, but it can translate into greater detail and greater understanding down the execution journey and expedite that entire process.

Cheryl Boehm: Yeah, and you touched on something interesting, like what if this had been around, you know, however many years ago, you know, just a few.

Tessa Burg: Yeah.

Cheryl Boehm:  But our careers have changed, marketing has changed so much with AI in just the last few years. So what does the future look like? Like how are our roles as marketers going to evolve as we continue to launch more agents and continue to increase this technology and AI into the future?

Tessa Burg: Yeah, I think right now, the only thing we know is that our roles are definitely going to change. And very similar to how our use of the agents and the feedback we’re getting is powering our innovation roadmap. It’s the same with how our roles will change, our use of it, and what we see ourselves doing that is positive, and having productive outcomes will inform that those are the types of things we’re gonna continue to do. So for example, as I use brand agent and more of the things that I would’ve manually done are automated, that inspires other agents, and now I’m going to increase my mastery of how to use these other agents, but it’s informed first and foremost by needs. The big question is, did I need new skills in order to use agent one, and do I need new skills in order to use agent two? And what we have found is the answer is yes. When we first started the AI Council, we had people raise their hand and they had already started themselves on basic training and prompts, but without a true responsible use policy, without creating trust and transparency, we weren’t able to get our use of AI scaled. And what we heard back from the staff is these general trainings simply do not work for everyone. So how our roles will evolve is going to be the cycle of I receive technology, so let’s say I am in the creative department, I receive the brand agent, and I also receive along with it the process and the expected outcomes for how that will be used. And then a skill building exercise. So here are the skills that you’ll want to build in order to use this agent most productively. And we don’t want skill building to take like months and years. And this is the other part I love about agents is this can be a very short training, but what’s involved in it would probably be, you know, reinforcing the right way to prompt, reinforcing the right way to use the output. And also then reinforcing our responsible use. How can we check for hallucinations? How can we check to make for bias? What are the points of validation that we need to see in order to know that what we’re putting into the next agent is accurate before we start to stack automations on top of each other. So those are all really important. And what I think that will lead to, if we think about that cycle of, I get access, I learn new skills on how to use this, I find points of validation, I start to think critically about how this aligns with the client’s overall goals, their brand, what we want for the client. And that builds and builds until what I foresee is a hybrid workforce of marketers directing multiple agents. And just like those agents inspire other agents, our use of them will inspire greater strategic thinking. The downside is, and I don’t know if it’s a downside, and this is a part, you know, we internally are still thinking about if I have more time to think strategically to solve bigger problems, the team is more strategic, it’s more senior, and it’s smaller. Like when I was giving those examples before, you know, those were junior, more junior level people who were helping us get all that data to compile it, to do the reporting, and now I’m doing that all, but I’m doing it with agents, and it’s automated and it’s in seconds. So that is a challenge that we’re looking at going into next year. But what I think is most responsible and what’s most important for marketers right now is building a platform that helps employees learn specific skills that elevate their strategic thinking and their creative problem solving. And just, I think that the more we collaborate on that together, then how we begin to change the roles for more junior level people will also evolve. And I’ll say we’ve had people at every level in the agency involved in the AI Council, and what hasn’t happened is that anyone’s been left behind or let go. You know, what has happened is everyone has elevated. So I think as we become more agentic, there will be new and different roles that people will grow into, and the more we continue down role specific department specific skill training, the higher quality of the work, the better the results, the better the margins, which opens up our ability to invest in innovation. So I could see those junior level, some junior level roles moving into this innovation space, because the execution space doesn’t require as much time and manual kind of tactical tasks.

Cheryl Boehm: Yeah, and that’s one of the things I love most about AI is that it does free up some time from those mundane tasks that no one wants to do anyway. And it gives you more freedom to think creatively, strategically, and spend the time and the effort on the things that we really wanna be doing anyway. So that’s exciting about AI and these agents as we continue to develop them.

Tessa Burg: Yeah, and I’ll say something else that’s come up is because we all have access to this technology, you know, there’s this thought that, and it has happened and then it’s come back, but that clients would pull this work in house, you know, like, well, I can create this agent. And what I’ll say is one is it’s a massive investment, right? Like we’ve been at this since 2016. We created our own proprietary LMS-like platform to help manage this transformational change. We have folks from Microsoft, and Standard Rapports, and Dun & Bradstreet, and software startups who have helped us think through a data strategy, how to leverage our data. And now because of that investment, that time, that expertise, it feels like we’re moving really quickly, but like it was a long road to get there. Do we think that clients should start evolving their marketing departments? Absolutely. But we’ve seen a few knee jerk reactions where we’re some will like pull it in immediately, but then it comes back, because there’s always the, well, there’s a few things. There’s the moonshot effect where they just over featured what they’re trying to build. The other thing is lack of understanding or just not being able to see how much investment is made up front. And so we’re sort of leaning into giving our platform that we’re using for upskilling and reskilling for marketers on the client side. You know, I’ve worked on the client side. I know it’s a very different environment than an agency environment. And when you’re on the client side, you have to be hyper-focused on the brand, the products, the roadmaps behind those things. And sometimes, it felt like marketing, though I knew getting it into market was incredibly important was 10 to maybe 20% of my day. So if all of a sudden, I would’ve been told while I was on the client side, we’re gonna pull it in 100%, well, now, I’ve taken my focus off my, the product, the brand, the solution using the customer feedback on the product and the brand to inform my company’s roadmap and collaborate with the product teams. So it’s, yeah, I’m just throwing that as a cautionary tale and I’m, you know, we won’t name names or anything, but like, you know, we’ve seen it. And so instead of waiting to see if that should happen, we want to help our clients. We’re offering AI audits now. Where are you? What things are you already doing as a client marketing team? What do they wish they could already do? Sort of get ahead of what might be like a board or a top down mandate like, “Hey, you should be able to handle all this right now.” Yeah, people should be able to do lots of things, but it doesn’t change the fact that it’s an investment. It also doesn’t change that agencies still have that advantage. And why I love working at the agency side of having already experienced things across many clients within same industries, within different industries to help expedite the application of best practices to accelerate a brand’s performance. So we know our clients have some really big things on their plate and if we can help with doing AI audits and giving them their own roadmap on what have they always wanted to do, what’s the best tech that they should be using internally, and then extend our innovation to their environment to accelerate their own roadmap, that’s something I’m really excited about doing. And we’ve just started and it gives those brands an advantage because a lot of businesses, again, I’ll say it again, 74% have been able to show value, still are trying to figure out how to even scale AI. And if you’re trying to take on more while you’re also trying to scale, that’s just making your road a lot longer.

Cheryl Boehm: Absolutely. Absolutely. So this has been a great conversation. I feel like we could talk about this all day, but unfortunately, we are out of time. So I know you can be reached on LinkedIn and as well as email if anyone has any questions and wants to talk more about agents, as well as the audits that we provide. And if you wanna hear more episodes from Leader Generation, just go to modop.com/podcast. That’s modop.com/podcast. And if there’s any other topics that you really want to hear about or have us cover, we also have an email address, you can email us there. And that is [email protected]. So until next time, we’ll see you then.

Tessa Burg: Thank you, Cheryl.