The peaks and pitfalls of AI innovation with Paul Roetzer

Speaker 1 (00:01):

It doesn’t matter what kind of agency you run. Traditional digital media buying, web dev, p r r brand, whatever your focus, you still need to run a profitable business. The Build A Better Agency podcast presented by a white label IQ, will expose you to the best practices that drive growth, client and employee retention and profitability. Bringing his 25 plus years of experience as both an agency owner and agency consultant, please welcome your host, drew McClellan.

Speaker 2 (00:34):

Hey everybody. Drew McClellan here with another episode of Build a Better Agency. Welcome back. Glad to have you. Uh, as you might imagine, everybody on the planet is talking about ai. We’ve already had several podcasts about it over the course of the years and this year, but, uh, we’re gonna talk about it again with one of the most, I think, well-known subject matter experts when it comes to marketing and ai. Paul Reer, who as you know, probably, uh, runs the AI Marketing Institute. Uh, he has an event coming up in a couple weeks that we want to tell you about and just in general, this, he’s just in the know on all of this stuff. And so I wanna get to that interview in just a second. But first I wanna remind all of you that we talk about a lot of things that we do for free webinars and all kinds of other things in our newsletter.

Speaker 2 (01:26):

If you are not getting our newsletter every week, they come out every Wednesday. If you are not receiving that, then you need to sign up. The way to do that is to go over to the agency management institute.com website, and you can do it one of two ways. You can scroll down to the bottom, and in the footer you’re gonna see a link for our newsletter. Or you can just go to the search icon in the upper right corner and type newsletter sign in form. I’d give you the url, but it’s weird with dashes and other things, so you won’t remember it anyway. But anyway, either way, you can go and sign up for the newsletter, and every week you will know what we have going on. Uh, we share links to the new videos we talk about, uh, you know, we’re doing the AI usage webinar series, which is free. We’ve got all kinds of other things going on, so worth your while to be a subscriber to that, I think. Uh, so go ahead and give us your name and email address and we’ll start getting it in your inbox. All right. So without further ado, uh, I want to get to the conversation with Paul, cause I have a million questions for him and I know you do too. So let’s get to it. Paul, welcome back to the podcast.

Speaker 3 (02:34):

It’s our annual AI adventure. <laugh>.

Speaker 2 (02:37):

That’s right.

Speaker 3 (02:38):

It’s good to be back.

Speaker 2 (02:39):

I’m, I, I suspect you’re just sitting around twiddling your thumbs. You got AI figured out and you’re just writing your next book and gosh, going on. Gosh, vacations. Right?

Speaker 3 (02:47):

I’ll tell you, we’re doing that weekly podcast and you know, we took a week off, uh, for travel, and you come back and there’s like 20 topics to hit it. It’s amazing how fast everything is moving, and I know every day there’s just something new and relevant and yeah, just trying to piece it all together, <laugh>.

Speaker 2 (03:06):

Yeah. I don’t, I don’t know how you and your team are even keeping current. I mean, I, everything changes every single day.

Speaker 3 (03:13):

Yeah. It, it’s, you know, it is challenging and oddly enough, we’re not actually using AI to do it

Speaker 2 (03:19):

<laugh>. So it is ironic.

Speaker 3 (03:21):

Yeah, it seems like a natural way. But, you know, I think the way that we do it right now, I think our weekly podcast became a forcing function for me and for Mike, my co-host, to not only stay on top of the news, but to actually synthesize it. And so what we do throughout the week is I have a very highly curated Twitter feed of AI people that I follow. We have publications we follow, and we just have a Zoom thread throughout the week as a sandbox for each episode. And we just drop the links and the things we’re reading and seeing. And then Sunday night we go through and we curate what are the ones that are gonna make the cut for the podcast. So like, we’re just reading and consuming all this stuff and then highlighting it. I have actually tested Chad Sheet, G P T and said like, what are the big AI items for the week that, you know, I might have missed? And it will do it. But that’s, we don’t actually use that workflow for the podcast.

Speaker 2 (04:10):

Right. So, um, for anybody who’s not familiar with you, just a very brief introduction, former agency guy, tell, tell this story briefly and then let’s talk about the conference and then what you’re seeing on the landscape for agencies.

Speaker 3 (04:24):

Yeah. I started my agency PR 2020 in November of 2005. We became HubSpot’s first partner agency back in 2007, wrote the Marketing Agency Blueprint in 2011. That was the year I started researching AI out of curiosity. Um, right. That eventually became my obsession formed Marketing Institute in 2016. Split it from my agency in 2018 2019. And then I sold my agency in 2021. And I’ve just been focused on the Marketing Eye Institute ever since. So the institute itself is seven years old. We’ve been talking about and researching and speaking about AI for a really long time. And then Chad t showed up and everybody started listening.

Speaker 2 (05:06):

Yeah. You know, wasn’t it crazy? It was like, no, it was like, it was not on the radar screen. And it was like within a two week period of time, everyone on the planet discovered chat G P T, and it was common conversation. It was strange.

Speaker 3 (05:19):

Yeah. It was November 30th is when it came out. And by the middle of November you could just feel the difference because I was getting phone calls from Yeah. A lot of people that hadn’t been listening or paying attention deeply yet. But I mean, big companies, big venture capital firms, universities, like everyone all of a sudden realized that this was going to change everything and they had to figure it out really fast. Right.

Speaker 2 (05:48):

Yeah. It was fascinating to me how it just, it was like somebody flipped a switch. Yeah. I mean you, you’re on what year four or so of your event? So it’s not, it’s not like this is brand new and yet, and I been on your podcast, it just felt brand new.

Speaker 3 (06:02):

Right? I mean, how many times have you, and I had this conversation, right? This is probably the third or fourth time over the last six years. You had me keynote your event two years ago, three years ago. Like it’s been around and people have been listening and people like you have been infusing it into your audience and trying to get people thinking about it. But yeah, it just sort of happened and I mean, 2019 when we had the first marketing AI conference, it was 300 people from 12 countries. And at the time we had, I don’t know, about 25 or 30,000 subscribers to the institute. So again, it’s not like no one was listening. Right. It just, it wasn’t, it was like an awareness thing. Yeah. They weren’t activating it really at scale. And I think that’s what changed is generative ai when it showed up, when Dolly showed up last year and then all these language generation tools, it just became so accessible to everyone. Mm-hmm. <affirmative> that like the c e o of a big company could go play around with chat G P T and be like, oh my gosh. And then walk into the marketing department and say, what are we doing with chat G P t <laugh>? Right. And they’re like, how do you know about chat g pt? That’s what happened is everyone could now use ai. And so it just became an urgency to solve for it.

Speaker 2 (07:10):

Yeah. I think you’re right. I think before it was, it was there, but it was theoretical, like, how will we use this someday? Or it was harder to see, see this few people using it. Yeah. Like I remember the first, the conference that you had and walking around and talking to all the exhibitors, which were amazing. And it was like, okay, this is like the beginning of how people might use this down the road. And then it, this year it seems like everybody is using it in some way, shape or form, even though we know we’ve been using it, you know, through other apps and other things for a long time. But it just now feels very real all of a sudden.

Speaker 3 (07:45):

Yeah. And you know, I think we had the AI score tool back in 2018. I built that with like 50 use cases for ai. Cuz a lot of what we were doing was trying to help marketers understand what was possible and agencies. Right. Right. And so you, you know, we built all these use cases and you could rate ’em and everything. They’re featured in the book, like the use cases were there, but they weren’t go go to chat g p T and just type in a prompt and you get a result. Like, it, it, the use cases weren’t that simple. Right. And it took like a, there was friction in the process to actually experience ai, but now image generation, video generation, language generation, music generation, like you can just do it all in many times for free. Right. And that’s, it just changed everything.

Speaker 2 (08:29):

Yeah. Yeah. So let’s, uh, let’s briefly talk about the conference and then I want to, I want to just hear from you what you’re seeing on the agency landscape. Cuz I know a lot of your community or agency folks and Yeah. And helping us sort of figure out how do we stay on the leading edge as our clients are looking to us to, to figure it out. So first let’s talk about the conference. Cause it’s coming up pretty soon.

Speaker 3 (08:52):

Yeah. The conference is, so we’ve got, you know, last year we had 200 people. We were still, as you know, as an event organizer yourself, still kind of in the midst of, are we traveling? Are we not traveling? Um, are we going to live events, are we not? Right. Right. And so last year was rough as a Yeah. You know, event owner. Uh, so this year we weren’t sure. We, we thought well, okay, well maybe we could double it. Maybe we could get to 400 people. And we surpassed that a while ago. Uh, so we, you know, surpassed the 400 goal. We expanded the space. So it, it’s live events seem to be back. Um, for sure. For sure. And then, uh, the interest in AI is obviously at an all time high. So this year looks great. Uh, the way we do the event is, you know, we look at use cases and technologies and case studies, the obvious stuff.

Speaker 3 (09:35):

But I’m a really big proponent of the need to do AI in a very responsible way. Mm-hmm. <affirmative> that augments, uh, humans. So a very human-centered approach to the application of ai. So we have ethics and, and and, uh, law and regulations around it, building responsible AI principles. Um, there’s a lot about how to really extend human capabilities and enrich lives and careers, hopefully with ai. So the conference really delves into a lot of that as well. So we’ve got almost, I think 30 sessions, uh, probably 40 different speakers looking at the current state of AI as well as where it’s going in the future. And then trying to piece this all together. There is a panel on, uh, specific to agencies, um, and then we have our separate AI for agencies summit we’re gonna do in the fall that’s gonna be dedicated to try and solve for this stuff with agencies. Cuz it is, it is a, a fair part of our audience and sure there are lots of threats to the agency right now, but there’s maybe like unparalleled opportunities. Um, I I think about it back when we became HubSpot’s first partner and I look at the opportunity then within inbound marketing and HubSpot and social media and blogging and all the things that were bubbling up at that time in 2011. And I think AI is 10 times. Oh gosh. Yeah. That opportunity at least like so much bigger for agencies.

Speaker 2 (10:57):

But yeah, I think you’re right. I think a lot of agencies, they are naturally curious about it. I think they have gotten past the point of head in the sand about it. They know that they can’t avoid it anymore. But you have a, you still have a lot of agency, especially agency employees who are feeling very threatened by what AI can do. And so, you know, I think, I think you’re right. I think helping agencies figure out how do I harness this? How do I, how do I use it to the benefit of both the agency and our clients, um, and, and still provide the best of what agencies do, um, that can’t be replicated, you know, by a machine. So I I I think this is a challenging time for agencies. A lot of people are feeling very threatened.

Speaker 3 (11:40):

Yeah. I, part of the challenge I’m seeing now is that the clients are becoming wise to what’s going on. Mm. So prior to the end of last year, you know, if agencies were looking into this, their clients probably weren’t asking for it. Right. Like, that was my experience of running the agency. We were building AI services back in 2018, 1920. None of our clients cared. Like, it just didn’t. Right. It wasn’t what they were demanding at all. So we were just too, too far ahead of the curve. Um, and now what I’m hearing is, you know, talking with some of these, these companies, they’re questioning the relationship they have with agencies. It’s like, well why am I still paying them X per hour? Like, they’re probably using these AI tools and doing things in, you know, a quarter of the time that they used to take.

Speaker 3 (12:25):

Um, there’s the issues of agencies that are racing head using these tools and maybe not realizing the issues with copyright around the stuff they create for clients. So there’s just, there’s all these challenges, but we now have a client base that knows these tools exist, knows that you should probably be using them as an agency and that may affect how they think about your billing structure and the services you offer. But we also have the opportunity side is most of these companies don’t know what to do. Right. They don’t know how to build a, a plan for ai, um, how to educate their team, how it affects their talent and tech stack. Like they’re, they’re, they’re wondering this and trust me, like I’ve talked with some Fortune 100 companies recently that are still trying to solve for what exactly this means.

Speaker 2 (13:12):

Yeah. Yeah. Well, and I think, uh, we just had, uh, Sharon Toric, uh, on the show probably, I don’t know, a month ago just talking about the, the copyright issues and the fact that, you know, all of this is of course, untested and, you know, people sort of struggling with legal language and how do we, how do we disclose to clients that we’re using AI tools and who does own what we create when we co-create through ai. So it, I think there’s just a lot of uncertainty right now around all of it.

Speaker 3 (13:42):

Yeah. And Sharon’s doing a session on kind of the current state of copyright and trademark and intellectual property at, at mayon for us. Um, because again, that was just one of those ones that it is a rapidly evolving space. I mean, OpenAI just last week there was a class, another class action lawsuit, I think against open for the training data that it appears it was trained on novels. It didn’t have in theory the rights to be trained on, cuz you can ask it to, you know, write based on a book and it’ll write in the same style as the author. So obviously it learned from that book. So there’s so many unknowns about this space moving forward. And I think that in a lot of cases for agencies, it’s gonna be best to take a pretty conservative approach to how you use these tools and what you pass through work for hire agreement as your actual work that they can have copyrights to. But that’s where I always say, go talk to Sharon, go talk to an IP attorney, like talk to someone who knows this stuff. But it is definitely a moving space right now.

Speaker 2 (14:39):

Yeah. So I, I think for a lot of agencies they are feeling threatened. So let’s talk a little bit about what you see are the threats. What are the realistic things that agencies should be worried about? It comes to ai. You mentioned a little bit sort of the billable hours and the Yeah. You know, I, I know I used to pay you $200 an hour to do this thing, but I’m betting you’re running it through chat G p t or something and it’s taking you five minutes. What are you worried about on behalf of your agency clients in terms of how it’s gonna change the business model?

Speaker 3 (15:10):

Yeah, you, I mean, you and I have talked about the financial models of agencies many times mm-hmm. <affirmative> in the past. I, we probably had a whole episode dedicated to it, but I mean, my, the chap chapter, one of my marketing agency blueprint book in 2011 was eliminate billable hours. So I’ve always been a, a huge proponent that, you know, agencies should be on some value-based pricing model. Right?

Speaker 2 (15:30):

Right.

Speaker 3 (15:30):

Um, so I think now more than ever, that is critical. So if you are an agency that is still using billers, I, I, I just think the end is near, like, there, there’s so many things you do that AI is going to assist in doing and make you more efficient at, which means you have to do more work or you have to flip it and just charge for the value of the outputs. Uh, so that to me is the biggest thing. And then I, I think the service side, you have clients, you have brand side people who are looking for help mm-hmm. <affirmative>. And if their current agencies can’t help them solve this, they’re gonna, they’re gonna have to go somewhere that can. And so whether you’re an ad agency doing advertising, creative and and budget management, the same way you’ve always done it and you’re not injecting smarter tools, the client’s gonna say, isn’t there a smarter way to do this?

Speaker 3 (16:22):

Or there tools we can be using. Right. If you’re a content creation agency, the same thing. They’re saying, well, are you using these tools? How are you using them? How’s this affecting your workflows? How’s it affecting your team? Um, they’re just gonna realize there’s smarter ways to do everything. Right. And so you have, you have to be the one that’s bringing those ideas to them. And that’s the main thing I say is just like, find someone or find a small team in your agency who can own figuring this stuff out and then staying at the forefront of it as it’s evolving and you just have to become more dynamic in your services and capabilities and be more proactive in bringing these ideas and the knowledge to your clients and creating value for them.

Speaker 2 (17:01):

Yeah. A lot of our agencies have put it together, basically a task force or a across, you know, a multi sort of disciplinary team of people who are experimenting and playing with different tools and then bringing it back to the agency to talk about, you know, this one has merit, or we’re gonna play with this one for a month. And, and, and interestingly, I think that the other thing agencies are doing is they’re saying to their clients, you know what? We wanna stay on the front end of this and we wanna do some experiments. Do you want, do you want to co-create some experiments with us? So I think, I think again, and you and I have seen this for years, I think, I think it used to be the model for agencies was you wanted to be the keeper of the knowledge and you sort of wanted to, you know, go behind the curtain, do your magic and come out. And I think the more sort of current model and AI is certainly part of this, is agency saying to clients, we don’t have this figured all out yet. There’s is things are changing too fast, but we do wanna experiment with it. We wanna stay smart about it. Do you want to do that together?

Speaker 3 (18:05):

Yep. Yeah. And I think you can, you can do it pretty quickly. Like it may seem daunting to figure all this stuff out, but I just go back to like, what is your area of expertise? Because AI is gonna affect every aspect of knowledge work, not just marketing and leaders agencies, but everywhere where we think we’re create or make decisions, like AI is gonna be infused into all of that. So if you’re an ad agency, like focus on smarter ad tools, if you’re a content agency, smarter content tools, email marketing, social media, whatever it is, just find smarter solutions. Always be testing, always be sharing with your clients what you’re learning, how you’re exploring it, and then you have to kind of figure out how to fit in the operational model, the financial model behind that. But I think you have to solve for innovation and value creation first. Mm-hmm. <affirmative> and figure out what that means to your agency and how it affects the staffing and the tech stack. But I just don’t see it being an option to not dive into this stuff because it, you could get left behind really fast, like Right, because it, like you said, it’s moving so quickly, the agencies that do jump in and figure it out are gonna be able to just create so much more value than the agencies that don’t.

Speaker 2 (19:15):

Well, and I think it’s, I think it was, again, it was speculative before, but now I think it’s very quickly becoming an expectation of we expect you an agency to get this stuff.

Speaker 3 (19:28):

I’ll give you an example. Like, this is, I haven’t talked about this yet, I’ll probably talk about it at maycom, but we needed to write the video script for the opening trailer for our conference. So, uh, Tracy, my COO, who runs our event, you know, message with me, she’s like, Hey, I need that script. So I, I wrote it over, I wrote the description for my keynote, um, which is gonna be a state of AI for marketing and business. And after I wrote that, I was like, you know, this is actually pretty good. Like this could work as the script if we expanded it out. So I sent it to Tracy over the 4th of July weekend. I was like, here you go. Like, if this looks cool for you, and by the way, I had this idea, like, we could turn this into the script if we expand it out, whatever.

Speaker 3 (20:07):

She comes back to me the next day, she goes here, I ran this through Chet g p t 3.5, and it wrote the script for me. What do you think? And I was like, I don’t know how to make that better. Like, it wrote the script in five seconds for the video trailer. And I realized like, wow, that that is what we’re talking about here. But I would’ve had to have spent probably three to four hours writing the script out, or we would’ve had to outsource it. If you’re, you know, on the client side you’re thinking about this. So we took a process that probably would’ve been five to 10 hours and we did it in five minutes. And then we’re using AI to actually build out the video itself to where something that you, we, you know, again, if I was a client or an event organizer, I think the y first year we paid like $5,000 to create the trailer, a minute and a half trailer for it. We’re gonna do this in-house in like probably less than an hour of time for no additional cost <laugh> using the free version of Chad g p t. That’s the future. Like that, that’s when I think about like every use case you can think about this and say, okay, as an agency, what are we even gonna need to do? What are clients gonna do themselves? Where do we move to? What remains a needed service six months, 12 months from now?

Speaker 2 (21:17):

Well, the interesting thing is, I, I do think a lot of agencies over the course of Covid and all of that, you know, have begun to shift from producing things and really spending much more time on the strategy side of the work and the persona building and the journey mapping and all of that. All of which I’m sure there are AI tools to help with. But yeah, it is the human insight and capital I think that agencies are bringing to the party. And so I think, I think agencies will begin to figure out how to find that balance even more so when, you know, it used to be that a freelancer could eat your lunch. Now Chad, g p t can do what a freelancer or your in-house writer would do in a couple hours, in five minutes. So the deliverable side of the business is gonna change pretty dramatically, I think.

Speaker 3 (22:06):

Yeah, I agree. And I, I do think that there’s gonna be a massive opportunity here to know how to utilize these tools in this way because, you know, it’s, it’s, it’s just technology. At the end of the day, it’s just smarter tech. And as agencies, we’ve been paid to use technology. Like if you think about HubSpot, that’s why I built an agency around the client could have learned HubSpot. Like there, there was nothing stopping them from hiring five people and running HubSpot themselves. They didn’t want to, right. They didn’t have the, the desire to do it. They didn’t have the, the budget for the overhead, whatever it was. And so at the end of the day, like clients are gonna pay you to figure out how to use these tools. So if you like a really practical example, if I want to infuse an AI writing tool across marketing, sales, and service, my two options are going to be, well, three are going to be, um, go get a foundation model like cohere or an, or open AI and build a custom model internally.

Speaker 3 (22:59):

Uh, the second is gonna be if Microsoft and Google Land AI built into three six, Microsoft 365 and Google Workspace, which they’re going to be launching and it’s really good, then you can just use the native features there. The third is you go get like a Jasper or a writer that’s an application or a software built to do this. Well those are enterprise solutions that need to be custom built. They need to be trained on brand style guidelines. They need to be prompt libraries built out. There’s entire service suites to be built on top of those solutions. Yeah. And so if I was a content agent right now, that’s probably what I would be doing. I would be figuring out which, you know, software companies I’m betting on, I would be infusing services to help clients ramp up very quickly on these writing tool capabilities.

Speaker 2 (23:44):

Yeah, I I think you’re right. I think, I think becoming the guide through and to the right tools and how to use the tools, and to your point, you know, I think ever since there’s been technology ever since there were computers, clients could have done all of that by themselves. Yeah. And some do, some do build in internal agencies inside their organizations and all of that, but a lot of them are like, you know what, I would, I would rather have that be a variable cost outside of my fixed budget. And so I’m gonna, I’m gonna hire experts. You spend a lot of time with agencies, but you also spend a lot of time with folks on the client side. What, what are they anxious about? What are they asking you about that agencies could be a solution for or have answers to?

Speaker 3 (24:29):

They, they have no idea what to do, right? Like they don’t have plans. So they know the tech exists. But again, we have to, you have to think about this in context of how fresh this is. So chat, C B T comes out in November 30th of last year. Then you have like the, you know, Christmas holiday, you’ve got, you know, everybody’s kind of off usually in December. So January shows up and everybody’s like, oh, this seems like a real deal by the end of January. You’ve got a hundred million users of CHE G P T by February. You have CEOs and boards and investors now saying, what are we doing with Chad g p T? Right? And by March, April, people are trying to figure out what do we do with chat G P T? Not what do we do with AI as a whole, but like this one specific thing.