Marketing Artificial Intelligence draws on years of research and dozens of interviews with AI marketers, executives, engineers, and entrepreneurs.
Marketing AI Institute Founder and CEO Paul Roetzer and Chief Content Officer Mike Kaput present the current potential of AI, as well as a glimpse into a near future in which marketers and machines work seamlessly to run personalized campaigns of unprecedented complexity with unimaginable simplicity.
As Mike and Paul approach the June 28 publishing date of their new book, Marketing Artificial Intelligence, they walk through the book writing process, as well as a detailed chapter-by-chapter explanation on their systems and thoughts.
So, come along on a journey of exploration and enlightenment. Marketing Artificial Intelligence is the blueprint for understanding and applying AI, giving you just the edge in your career you’ve been waiting for. Learn more from Paul and Mike:
[00:04:18] A framework of the book
[00:05:59] Creating a resource center for the co-authors
[00:09:37] Leaning on our own performance data
[00:12:21] Chapter 1: connecting the dots
[00:23:50] Marketer + Machine scale explained
[00:28:26] Getting started
[00:33:55] AI + marketing disciplines explained
[00:51:17] We want to hear your stories
Disclaimer: This transcription was written by AI, thanks to Descript, and has not been edited for content.
[00:00:00] Paul Roetzer: Welcome to The Marketing AI Show. The podcast that helps your business grow smarter by making artificial intelligence approachable and actionable. You'll hear from top authors, entrepreneurs, researchers, and executives, as they share case studies, strategies, and technologies that have the power to transform your business and your career.
[00:00:20] My name is Paul Roetzer. I'm the founder of Marketing AI Institute. And I'm your host.
[00:00:28] Welcome to The Marketing AI Show. We are back for episode 19. This is an ad free episode. this says special book launch edition for our new book, marketing, artificial intelligence, AI marketing, and the future of business, depending on when you are listening to this, it is available, uh, now or very shortly and June 28th, 2022.
[00:00:53] It comes out in audio. Digital and print. It is with Matt Holt, Matt Holt Books, such as an imprint of BenBella books. Uh, Matt is a, a longtime friend of mine. He was my former publisher at Wiley for my first two books. Um, so I'm, I'm super excited to be working with Matt again. He's incredible. His team's been awesome.
[00:01:11] So, uh, I am Paul Roetzer, founder and CEO of Marketing AI Institute, coauthor of Marketing Artificial Intelligence. I am here today with Mike Kaput, our chief content officer and my co-author what's up, Mike?
[00:01:24] Mike Kaput: Not much. How's it going?
[00:01:26] Paul Roetzer: We're launching on a book next week, dude. I know. I know. So it's a little, it's weird feeling.
[00:01:31] It really is. It's so funny because, and I think like part of the idea with this podcast, we're gonna give you like an overview of the book, give you some insights from the book. Um, but also a little bit. Background the process. If you haven't written a book, um, you know, share a little bit of what goes into it these days, uh, specifically talk a little bit about how we used AI in the book writing process and some of the ways we didn't, but that we thought about using.
[00:01:55] So if you listen to our AI for event marketing episode, it's kind of in a similar vein where. We had all kinds of cool ideas of how to use AI, but we didn't get to pull it all off. But I definitely think there's an opportunity in the publishing industry to shake things up with smarter approaches, to, to, to content creation, to publishing, to writing.
[00:02:15] So we're gonna give a little bit of, of insights. It, this book is how we, I started writing it. I remember I was on spring break, 2021 writing chapter one. So I, I, I started writing my portions of. April of last year and then rewriting all the way up until spring of this year and the way we did it. And, and again, we'll kind of walk through the book, but the book is 17 chapters.
[00:02:39] Um, we're gonna kind of give you an overview of each chapter, but the middle part of the book is this piloting AI section. We're actually go AI for. Advertising AI for community different categories of marketing. So Mike fo focused most of his writing in those middle chapters. Um, so that's kind of how we divvied it up and we, you know, tackled the middle part of the book.
[00:02:58] And then I sort of wrapped it with the, the, the front end story of how, how we got here, what the opportunity is, um, how to think about AI within your business, within your career, and then wrapped it at the end. With, what does it mean to you? And, and how do we think about this moving forward for good, like for good of society for good of business.
[00:03:18] So I think it came together really well. It was a pretty fluid process. I, I, I would say Mike, like, right. I mean, it was for sure it was interesting to do it and again, I'll drop in some things as we think of it. Like we did it in Google docs. Um, I'm a part of a, like a private Facebook group for authors.
[00:03:35] And I know there's like, what is that one scribble or scribe? What's the one, the writing platform that a lot of people like to use. Do you remember it? I think it's called. That sounds right? Yeah. So there's a bunch of writers, like swear by that we, we didn't use it. We used Google docs as is, and I, I found it to be ver very doable in Google docs.
[00:03:55] Um, one of the benefits to the Google docs is voice to text. So I would often be driving and think of something for a chapter. And I could just call up that chapter on the app, on my phone, and then. Easily like voice to text my notes for that chapter. So the way we did the book, again, this might be interesting to people it's relevant for content publishing as a whole, probably.
[00:04:18] Um, you can jump in here. Might you think about, but I, I remember we. First we went through and created the template for what each chapter would look like. And then in a Google drive doc, we created all the chapters in advance. So basically we built sandboxes because it wasn't this linear thing. We didn't write like chapter one through 17 in perfect order.
[00:04:39] Because what we did is we have published what, eight or 900 articles about AI. So we had a ton. To start from as a base, then we've done dozens of podcasts and webinars and videos and all of that audio, we converted into text. So we did, uh, speech to text transcription through otter.ai and the script.
[00:05:03] That's how we did that. And then we just took that stuff and we would like copy and paste and throw it into the potential chapter. So it was just like, almost like a brain dump of everything we had, which was probably, I don't know, like two, 300,000 words, probably like when it was all said and done. So for a 60,000 word manuscript, we started with hundreds of thousands of words.
[00:05:26] As a basis that in some cases we repurposed or rewrote, or just used as a reference like, well, how did we say this before I came across that a lot where it's like, I wanna say it differently. And I want to tell the evolved story, but how have I previously explained this topic to people? So. I would say the first couple months, at least for me was just that process, like the discovery on our own content, organizing our thoughts, nailing down the outline.
[00:05:55] Can you think of anything else we did just in the prep process? Um,
[00:05:59] I do think, and you kind of alluded to this. We did create kind of that overall resource center of just pulling everything together. And I think at like a really high level, what was cool about this process and showing people the behind the scenes is that while we did use AI for audio transcription and some other things we can talk about during the writing, um, the just planning and having some processes and having some systems and heavily.
[00:06:27] Leaning into repurposing or reusing content is just a smarter way. I think, to do a book, if you are creating content regularly, and I would argue that you probably don't need 800 plus articles to have a book in you. If you've been someone who is. Uh, you know, a content creator, um, for several years you probably
[00:06:49] have a book.
[00:06:50] Yeah. It's, it's a great point. And again, if, if you just create a bunch of social videos or you put a post on LinkedIn, a bunch, like just curate all that stuff and put it into a structure of a story. And I think, you know, this is an interesting thing somebody's asked me before recently on a podcast about.
[00:07:06] The book writing process. And like when did I decide to do a book? And I mean, the answer is actually I thought about doing this and I tried to do it in 2018. And what we realized at the time was I wasn't ready to tell the story. I didn't know the story. Like we. I almost felt like I knew the beginning in the end.
[00:07:24] Like I, I had a really good sense by 2018 of the story of AI through the decades and how it evolved and how it, we had arrived at this point where AI was now could be commercialized. It could be applied to marketing and business today in very practical ways. And I thought I knew where it was all going to go, where it was gonna take the business world and marketing, but we didn't have all the middle part.
[00:07:47] We didn't have all the use cases nailed down. The technology was so early, it was hard to find legit vendors to feature. There weren't brands we could find that were actually doing it. Like it was very, very early in 2018. Yeah. And so we shelved the idea of doing the book for a couple years because I, I just didn't feel like the market was there yet.
[00:08:06] Um, So that was, I don't know, that's kind of an interesting thing. So we just kept building the content and driving, you know, both growth, the audience. And I think one of the other ways we probably used AI indirectly is our own content strategy for the Institute that might yeah. Really drives. It's largely dictated by AI tools that tell us what to write and what will rank best and you know, where there's demand for information and things like that.
[00:08:31] And so that indirectly then affected which chapters of the book we wrote.
[00:08:36] Mike Kaput: Yeah, for sure. I think that's a huge point here is that, you know, we're using several AI tools and platforms to identify those opportunities and then understand exactly how to cover topics, to rank for them. And obviously the book is not designed as an SEO play, but if you think about it, We have all the data on which areas are most popular in terms of AI for different types of marketing.
[00:09:05] So while it doesn't really matter, if a chapter in the book. Optimized for search a that content can be repurposed on the website and rank really well. And that can drive further people to the actual book. Or also you can say, okay, these 10 categories are widely the most searched or the most, uh, the most, um, talked about online.
[00:09:30] And that can give you a really interesting idea of what the potential audience is within your broader.
[00:09:37] Paul Roetzer: Yeah. So for that piloting is section, we definitely leaned on our own performance data. We also used like N I CS code data for size of industries to help us get a, a sense of, you know, how big a market potential different stories we've had in terms of potential interest level.
[00:09:51] And we used LinkedIn sales navigator data to look at number of professionals and specific areas. So like we tried to layer in data wherever to inform some of the story. Honestly though, like the front and the back end, the ch a lot of the chapters I focused on was more just the story that I, I thought needed to be told.
[00:10:10] There wasn't performance data that was going saying, okay, the first chapter should be about this. It was just how we got where we are. Like, how did we go from Mike? And I are both just writers by trade. I came out of journalism school. Mike has a journalism background. And so in 2011, 2012, we started thinking.
[00:10:28] AI. And then really in 2016, we launched the marketing AI Institute. Um, we're just telling the story of like how we figured this stuff out because we're liberal arts people. We're not, you know, we're not data scientists and machine learning engineers. And our whole mission with the Institute has always been.
[00:10:44] How do we make this make sense for people so that they care so that they realize the opportunity that exists to grow their own career and to grow their business in a smarter way. And so we're always trying to like simplify the message and make it make sense. And so, you know, kind of jump ahead a little bit here.
[00:11:02] I think that that's what chapter one of the book is all about. So chapter one, Tells the story of the science of making machine smart is the title. But what really it's about is laying the foundation of what is AI, how is it a very practical thing, but then we tell the story of Microsoft, Amazon, and Google, and their efforts to infuse AI over the last two decades and how they've put hundreds of millions or billions of dollars.
[00:11:29] Into the development of advanced technology that we as consumers experience every day. And so that's kind of the story of that first chapter is this isn't new look at the biggest tech platforms in the world. And they believe that AI is not only the future. It is the present and they're doing everything in their power.
[00:11:48] To educate the market. Think about the Microsoft AI ads. You see Google AI has its own Twitter handle. AWS has dozens of pre-trained models. Like they're trying to make this stuff accessible because they see it as the future of business. And so we start the book off with this story, like, Hey, this isn't about just a bunch of use cases for marketing.
[00:12:07] This is a much bigger story. And the biggest brands in the world are driving innovation. That are going to alter your career path and your company's success, whether you embrace this stuff or not. Yeah. I, I really
[00:12:21] Mike Kaput: think. Coolest thing about the first chapter here that you wrote is just that connecting the dots in an extremely clear way.
[00:12:29] Like any marketer that reads this chapter is going to come away with a really solid understanding of how did we get here, but connecting the dots shows why it matters. And I think one of the points that kind of jumped out to me that I wanted to kind of get your take on is there's obviously some really important milestones along the way in the development of AI, but what.
[00:12:50] Really stands out as a high level point is that the rate of technological change is accelerating and AI is largely driving that. So we're not talking about, uh, you know, making a better widget. We're talking about exponential change, exponential acceleration, exponential growth. And that's why we will talk about this.
[00:13:12] Maybe a little later, we've seen in the last year. Some of the most mind blowing tools, uh, we've ever seen in AI, that seemed like they came outta nowhere because we're on an exponential development curve. And so do you, I wanted to kind of get your take on that just about the rate of change and why that matters for
[00:13:29] Paul Roetzer: marketers.
[00:13:30] Yeah. So the first chapter, you know, again, there's, when you write a book, I mean, collectively Mike and I probably spent five to 600 hours on this over a year and a half. Um, and I there's moments where you very specifically recall. Where you were and what you were doing when you wrote something. And, and chapter one to me is like probably the biggest of those, because I was in Florida spring break, 2021, um, with my family's staying at, you know, my, my wife's grandmother's house.
[00:13:58] And I. I had already written most of chapter one. And then I was reading genius makers on that trip by C Mets who, if anybody was at our Macon 2021 virtual event, Cade Mets was a, a keynote. We went through his book, genius makers and he tells the story in genius makers of deep learning, coming of age as the advanced form of AI, that really gives machines the ability to have language and vision and, and things like that.
[00:14:23] And, and really advanced prediction capabilities. and it was sort of that light bulb moment for me when I was like, oh my God, like, this is why I couldn't tell the story. In 2018, the, the tech, the AI actually actually hadn't advanced to the level. We thought it was at that point. And so genius makers tells this incredible story going back to like 2011 of the advancements of deep learning, specifically at Microsoft and Amazon and Google and deep mind and open AI and some of these other major Invidia and, you know, other companies.
[00:14:57] And that was the moment I was like, oh my God, I have to rewrite this whole chapter. Like the, the story I was gonna tell, just took out a whole different perspective. And I remember like standing there's a park across from my, my wife's grandma's house. I was standing at a garbage can. So my laptop could be at like standing desk level in a park, retyping this chapter with the story of genius makers running through my mind.
[00:15:20] So there's actually. Again, if you read it, sometimes it's cool to, to read a chapter and have the context of sort of where it comes from. And so I could like point out exactly like the paragraphs where I, I pivoted basically. And, um, had a different perspective on the story that again, that I would not have had in 2018.
[00:15:39] Mike Kaput: Yeah, for sure. I mean, it's just, and then, you know, as we're talking about how. This stuff, uh, comes of age and is developing. I mean, you mentioned Amazon, Google, Microsoft. I, I really think the chapter wraps up really interestingly and strongly, by, like you said, deep diving into what those companies are doing.
[00:15:58] We're not just mentioning them. There's pretty extensive sections on exactly how they're using it. And I really think you ju it just ends the chapter. If you're a marketer, a business person, you look at that and you're saying, oh my God, how do I start using this stuff? And that's why I kinda like, as a great transition between that chapter and chapter two, uh, you dive right into a framework to start thinking about this.
[00:16:22] So, uh, we call it language, vision prediction as the name of the chapter and broadly kind of the way to be thinking about this, but I loved how you started this off and. You know, one of the biggest challenges I faced as a marketer, trying to comprehend AI was finding practical applications and use cases that made the technology seem less abstract and more actionable.
[00:16:45] So really connecting the dots. Could you kind of explain how you got to that and wh how language, vision, and prediction come in.
[00:16:53] Paul Roetzer: Yeah. So this again, probably goes back to around that 2018, 19 timeframe. So that the Institute was a couple years old. We were launching the conference, the marketing AI conference, and I had this thought like, We're trying to make this tangible to people.
[00:17:10] How are the big companies positioning it? Like, what is Microsoft, Amazon, Google, how are they explaining AI solutions? And so that led me to just go to their sites and start looking at how are they categorizing the solutions they offer? How are they explaining it? How are they educating around AI? What started emerging was they all had different ways of describing it.
[00:17:31] Like, you know, for example, language and speech might have been differentiated in some of 'em and they've evolved since then. But what seemed uniform to me was there was three main categories of AI applications and that was language. Um, Which generally speaking is the understanding and generation of language.
[00:17:49] So Surry Alexa, their ability for to understand when you say something, quote, unquote, understand it's math. They're not truly understanding, but language understanding. And then generation is speaking or creating text. Like it's the generation of language. Now there's other versions like content, summarization, and text to speech and speech to text and translation.
[00:18:09] And those are all forms of language. AI. Came to kind of encapsulate it. Vision was another one, computer vision image, generation image recognition. Um, deep fakes is an example. You'll hear here. Uh, we may talk a little bit more about it later, but like Dolly two, as a technology that recently came out where you give it a text prompt and open AI platform will generate an image from pixel level up an original image.
[00:18:34] Um, Google released image in, uh, a month later in may of this year. Same idea. There's all kinds of crazy applications of VI vision technology. Um, and then the final is prediction. The biggest one, there is personalization as a category. So you're trying to predict human behavior and outcomes, and therefore you're able to personalize based on data.
[00:18:55] So language, vision, and prediction. As these kind of macro level categories is what emerged out of a study basically of the market and how it was being defined. And it gave us a framework to start educating people. And we use this all the time in our intro to AI, for marketers class that we teach every couple weeks.
[00:19:12] Um, and, and we show that and I think it helps people like, oh, okay, like this is actually understandable. This isn't that complicated. And so we follow that in the book and really unpack what each of those areas are and give you some very practical examples of them. Yeah, I
[00:19:27] Mike Kaput: love that. I feel like any marketer will go through that and come away.
[00:19:31] Just understanding how to start thinking about this technology in a very, very simple way. It is not rocket science to be able to go through these three areas and start thinking about how to use it in actual business use cases. One, uh, one thing I did wanna call out given kind of some recent developments in the world of.
[00:19:54] Uh, you did have a, a section here called how can AI inspire our creativity. And it's really about this balance between what does AI mean for human creativity? And I think there have been some developments in the last few months here that you mentioned, like Dolly too, that. Make this a really interesting question.
[00:20:16] Do you wanna kind of maybe talk through why you included
[00:20:18] Paul Roetzer: that? Yeah, so I, I felt like we had to address it in the book. Like what is creativity and can machines be creative? It's a very, um, big question in, in business and society. Um, certainly in our industry and you know, it's funny. I, I haven't gone back and reread that section since Dolly two came out two months ago.
[00:20:42] But I, you know, knowing what I wrote, I don't know that it would've changed what I wrote, cuz I think where I landed. AI can be creative. It just not in the way humans are creative. So I think I gave the example in the book of, um, a dog, like mm-hmm , you know, the AI doesn't know what a dog is. Like, it can create an image of a dog, but it has no experience with dogs.
[00:21:08] It has no emotion tied to dogs. It doesn't understand. Why humans love dogs. Um, there's nothing about the machine that understands human emotion and, and empathy and all the things that, you know, give us the ability to know what a dog actually is. The machine doesn't know what a dog actually is. It just knows that.
[00:21:28] The image looks like a million other things that it was told was a dog. And so whether it's writing poems or creating music, my whole point was like the way a human does. Those things is the sum of our experiences, our life experiences, the emotions we've felt, the things we've gone through. That's what comes out in our words that is not.
[00:21:49] A machine like it doesn't have those things. It has data that goes into it and it finds patterns and it creates outputs. So I think the key though is like, it, it doesn. At the end of the day, it doesn't matter. Like there's this big battle right now is, you know, Lambda Google's, um, language model. Is it sentient like some researcher at Google got put on leave?
[00:22:12] Cuz he made an argument that their language model had become sentient. I was aware of itself basically. I was conscious. It's not, but the point is an AI researcher at Google thought it was. And so my whole thing with creativity and Senti and all that. It doesn't actually have to get there or be truly creative for it to matter as long as it can generate creative outputs or it can convince people.
[00:22:39] That it is aware of itself. And so that's where I think this, this chapter is really important to give a broader perspective to people before we like really drill in out the use case and technologies and the things that you can just go do right now to create value with AI. I feel like it's really important that people understand the bigger picture of what is going on and what's happening.
[00:22:58] At these major tech companies that is going to influence like we're using this AI every day in Google and Netflix and Spotify and apple and TikTok and like LinkedIn, like it's everywhere. So whether you want to, or not, like you're already affected by AI and it's just gonna find its way into every piece of business and marketing software use as well.
[00:23:18] But I mean the creativity one, I, I could talk all day about, like, I think it's one of the most fascinating topics in AI, right?
[00:23:25] Mike Kaput: Yeah, for sure. Um, so I wanna make sure that we get to kind of how people can really start. Applying this stuff. Mm-hmm and I, I love again, just the way this is structured. And I really think this is a Testament to the process we talked about is really thinking through step by step, the pieces of this before even getting any of the content before even writing anything.
[00:23:50] So chapter three is really getting into starting to get into the, how mm-hmm . So you've learned how to think about this stuff now. How do you actually sit there and. AI technology. And this chapter is called the marketer to machine scale. Do you wanna, maybe just kind of briefly describe what that is and why it's
[00:24:11] Paul Roetzer: important.
[00:24:12] So if anybody hasn't seen it, we've talked about the machine market to machine scale before it was something we created. Um, probably back in 2020, I think. Uh, but the basic idea is to provide a rating scale of how intelligently automated an AI. Function is AI technology is, and we, we parallel it with the autonomous vehicle industry.
[00:24:33] So in the, in that chapter, I tell this story of Tesla and how there's a standard measurement within the automotive industry on, on zero to five as their scale of how autonomous a vehicle is. And the key aspect is what is the driver doing? And if the driver is in the car and required to. Oversee what the car does and be able to take control of it.
[00:24:55] It's not fully autonomous. So level five autonomy in the automotive industry is there's no steering wheel needed. Human just gets in, says, I want to go to Chicago. And the car takes them to Chicago. Human does nothing in the process. We don't have that today. Tesla, which is the most advanced we have is level two.
[00:25:14] The human still has to, you know, oversee. Starts to provide inputs, tell if it's doing the right thing or not. And so we've always wanted to have the same kind of thing for AI and marketing a, a scale that says how much does the human need to do? How involved is the marketer in using this technology? And the base idea we get to is.
[00:25:32] Most of the tech, you uses level zero, it's all human all the time. You're writing all the rules. You're telling the AI what to do, or you're telling the software what to do, send the email at this time to these people. Here's the subject line. Here's the body copy the machine's doing nothing on its own other than allowing you to send it.
[00:25:48] Whereas. As you move it to level one and level two. Now the, now the AI assists you. Maybe it tells you what segment to send it to her. It writes the subject line, or it personalizes send time to each person. It, it can make these, these incremental changes that make a big difference in your sending of the email level three and level four are.
[00:26:07] Level three's possible, but it would take a ton of training level four doesn't exist. So it basically gives you an over an overview and a framework to say, okay, so I don't need to go from zero to full autonomy. That's not what I'm trying to achieve here. As a marketer, I'm trying to find AI. That can assist me augment what I'm capable of doing.
[00:26:24] Make me more efficient, make me better at my job, create a greater probability that I'm gonna achieve my goals. That's what AI does. It's just smarter technology. And so that chapter really helps you visualize that idea and then changes your perspective as you go and start talking to these vendors or your existing tech stack, to be able to say, okay, how do I, how do I go from zero to one?
[00:26:45] How do I just, you know, save instead, I was doing 10 hours a month on this. How do I get it down to one? And then how do I redistribute those nine hours to something else? That's the basic premise of that chapter.
[00:26:55] Mike Kaput: Yeah. And what also struck me as really helpful here is you end the chapter with a number of questions that you need to be asking vendors and about the technology.
[00:27:07] So, you know, those are really helpful pieces of content, but why is it so important? To have these questions
[00:27:12] Paul Roetzer: in here. Yeah. So like we said, it's, it's just buying smarter technology, but it's really important that you understand how it's gonna affect your team, your organization, what kind of resources the, the vendor's going to provide, because it's like any other marketing technology, you have to be able to onboard it and then get the value out of it.
[00:27:31] You don't wanna buy this tech and then have it sitting there using 10% of what it's capable of. So you want to go in and know, okay. One is this legit AI technology, is it actually going to make us better, more efficient at our jobs? Does the company provide training and onboarding? That'll help my team learn how to use this new technology.
[00:27:50] Um, and then from there, you just start to think about, as we scale this up. How is it gonna affect our organization? So yeah, we give, I think there's about two dozen questions. Yeah. And they're very simple questions, but it helps you realize like the kinds of things you're gonna want to ask of these vendors that are claiming AI.
[00:28:06] Because the other thing we teach in there is like, not all AI is created equal just because they say they use AI or natural language processing and machine learning doesn't mean it's gonna make you better at your job, or be much smarter than what you're doing. So you don't need AI for AI's sake. You, you need smarter technology that makes you better at your job.
[00:28:26] Mike Kaput: Absolutely. And, you know, I think as we dive into chapter four, it's really very on the nose title that everyone is probably wondering as you listen to this, or as you read the first three chapters, how do I get started? And chapter four is called, getting started with marketing AI and quite literal . Yeah, I like it.
[00:28:45] um, the, this all kicks off with, um, the five PS of marketing AI do now. Before we dive into that. I think some of this really came about from. We did probably a hundred plus spotlights on different AI vendors over the years where we would send a common list of questions to vendors. We were interested in, in the marketing and sales, AI space.
[00:29:13] And over time we were able to help them tell, or they were able to tell the story of their solution. And we would ask things like, how does your solution use AI? How is it smarter than other technology? And out of this, I think we very quickly started to see. That the use cases and the vendors and what they were capable of, started fitting into a framework, um, within marketing specifically.
[00:29:39] And that's really, I think how the five PS started to form. Do you wanna kind of describe what the five PS marketing AI are?
[00:29:47] Paul Roetzer: Yeah, it, it you're you're right. That was kind of the process. And then when we were building AI score and trying to like categorize the use cases, what we were doing, it was like, okay, We knew like five examples in content marketing and three in advertising and two in communications and six in data.
[00:30:03] And so we could try and help people understand AI by showing these use cases by category, which is in, in essence, what we end up doing in the book, but to try and lump them all under more logical groupings. We thought more about, well, what do marketers need to do? So planning is the first one. So that could be pricing strategy.
[00:30:21] It could be content intelligence, it could be ad budget management. So planning is like a broad category of building intelligence strategies for your organization. Then you have production. So like creation of intelligent content. That could be an ad content, email content, you know, content, marketing, blogging, podcasting videos, whatever it is.
[00:30:39] So it's again, kind of like a broad category. Then personalization is powering the intelligent consumer experiences, taking the data. You have access to figuring out how to apply it to personalize, create convenience, drive behaviors, and outcomes that you desire. Promotion is the cross, uh, channel promotions of the content and ads you create and social, and then performance is taking the data and turning it into insights and actions.
[00:31:03] So we don't just have a bunch of data sources. So those are the five groups that we started then putting use cases under. And those use cases again, could come from any number of categories, but if you go through AI score, which is. score.marketing AI institute.com. We talked about it in the book, give some sample data from a survey we did with drift.
[00:31:22] Um, you can walk through and actually rate 50 use cases on the value to you to intelligently automate those tasks, which is, as you said, the getting started with marketing AI. The first way to do it is at a use case level. You, you find a bunch of things that you already spend time doing. You look for repetitive processes or data driven processes, and then you go find an AI tool that can help you do that thing.
[00:31:44] That's you could leave this podcast episode and just go do. I write blog posts all the time. I need to come up with headlines or, um, you know, I create ads, social ads, or I manage emails and, and you just go through and write a list of all the things you do. And then how many hours you spend doing. I mean, you just go find a AI tools to help you do them more effectively, um, saving yourself time and money.
[00:32:05] So that's those all kind of connect AI score. Parent categories. And so we really explain all that. And then the second part in that chapter is the problem based model where it's a bigger picture. It's like, okay, we are a media company. We are an event company. We are a dental practice. We are whatever, what are the big challenges we're having in our business?
[00:32:25] Where are the inefficiencies? Where are the where's the biggest values we can create? And then you go through a system. We give a 10 step, a 10 step F. Where you can more intelligently assess ways to, to, to achieve that goal, um, differently than you previously have by using AI technology. And so that chapter gives you that use case and the problem based framework, and then walks you through a bunch of examples and gives you access to a workbook you can download to, um, to do that use case model.
[00:32:55] Very simply.
[00:32:55] Mike Kaput: And we we've seen that workbook to be really popular and for good reason, I mean, look, there's a lot of interesting business marketing books out there, but I, I don't think as many as they should go into this kind of level of practical and actionable detail about how to actually do everything that we're talking about.
[00:33:13] It's important
[00:33:14] Paul Roetzer: to do. Yep. And then that leads us to the piloting AI chapters, which is again where Mike did most of the heavy lifting on those. Some based on content we'd previously created and kind of curating the most popular stuff. And then, you know, augmenting what we'd done before and some, you know, new chapters, but there's 10 chapters in the middle of the book.
[00:33:34] So why don't you talk to us a little bit about. The format of those chapters and sort of what people can expect, because I think we kind of designed it as like, almost a choose your own adventure in those chapters. Like you don't do eCommerce, don't read eCommerce chapter, like, you know, skip around, but you can kind of pick of those 10.
[00:33:49] So why don't you walk us through what those 10 are and then like the format of those chapters. Sure. So
[00:33:55] Mike Kaput: we chose 10 broad areas of marketing sales to look at how AI can be deployed, you know, starting today by professionals in these industries. Now I'll get into in a second how we picked them, but. To the point of choose your own adventure.
[00:34:11] I'm gonna read 'em off. In case someone listening, uh, has a particular interest in one of these. So we did advertising in AI analytics, communications slash PR content, marketing, customer service, eCommerce, email, marketing, sales, SEO, and social media. And each of these chapters is called category and AI.
[00:34:34] And a lot of this stemmed from what we had. Been writing about, been thinking about, and actually been piloting in the marketing AI Institute side of the business before we even started the book. In some way, as we are building the Institute, we have been trying to solve for all of these categories. At some point in the life of the business, we have done all these types of marketing and our hypothesis is AI can make every area of marketing smarter.
[00:35:06] So what these chapters actually do, they all follow the same format where we tee them up and say, look, here's real world use cases or stories of. AI is being used in this particular area of marketing and some of the pretty amazing results you can get with it. Then we dive into actual use cases for the technology at a high level.
[00:35:28] So for instance, within advertising, we're talking things like use cases like AI can do media planning for you. AI can help, uh, regulate budgets for ads. AI can help you create ads. So we go into all these very interesting. Use cases for the technology. And then we end each of these chapters by, uh, providing an extensive list of tools that we know of, that we have used ourselves in some cases that are highly recommended to go try and start.
[00:36:02] Doing these things and trying to start finding solutions that can help you do AI for whatever. So if you are in any of these areas, you come away very quickly with. A really clear picture of, oh my gosh. Like I can be using AI and advertising for a, B and C things. Here are tools that do those things. I've got all the frameworks and questions from the previ, the first four chapters.
[00:36:30] I am ready to go out and start. A free trial with a tool, a demo, or having a conversation with a vendor. So that was really the idea with this whole middle section is to get people actually taking their first step, whatever that may be, but actually taking action to get started with AI. And what was pretty cool is we actually did lean on.
[00:36:55] AI specifically content marketing AI to guide us on these. I mean, we had years of data from content we published on what people were most interested in from a category level. Not to mention we had. You know, AI powered cer uh, AI powered, um, SEO strategy that was telling us, okay, here are the other categories that get the most traffic are most interesting and are just most popular among audiences.
[00:37:26] So, you know, there are probably some categories in here that we did not have time to include, but all of these are areas that we found. We the ones people were very interested in, um, using data and using AI to kind of tell us stuff about that data.
[00:37:43] Paul Roetzer: Yeah. And a couple of notes. So there's more than I think.
[00:37:48] Total there's like 70 some vendors, but a couple are in there twice by by categories. So there's over 50 unique vendors that are featured in there with URL description. Um, so yeah, I mean, again, we designed this to be very actionable. So even if you just read the opening chapters and then you dive into like, okay, I do advertising, you could go and start applying AI right away.
[00:38:12] Um, so that was the one note I had the other is. What we've found is after six years of doing this, you have to make AI personal to people. Like you have to explain it in the context of their career or the things they do cuz AI on its own is kind of abstract and broad and can be scary or overwhelming.
[00:38:32] But when you say AI for advertising, it's like, oh, well I, I do advertising. Like I, you know, that that's interesting to me or AI for blogging or AI for podcasting or AI, for whatever, like. If you just make it to them or AI for healthcare marketing. So we, we have, as Mike was saying, like, we have this content strategy to make like AI and like AI four it's it's for this, this specific personas for, for this specific vertical, because we've found people are more likely to pay attention or to, you know, attend a webinar or adopt, um, or seek out knowledge to adopt AI.
[00:39:05] If you actually. Bring it down to, to where they're at in their learning journey in their career. Um, so that's really the idea behind these middle chapters. And then the other thing I'd add. Stay tuned to the podcast because we're planning on going through over the next, like, you know, 12 weeks or so. Uh, each of those chapters in depth and talking more and, you know, even beyond what's in the book like advertising and AI, we did one already.
[00:39:32] We're gonna go through each of those areas of marketing and really talk about and explore what's possible with AI and how it can again, build smarter career paths, smarter businesses, um, using. More advanced technology, basically. Yeah.
[00:39:47] Mike Kaput: And the last thing I would add here is if you are listening to this or, you know, reading further about the book and you're saying, whoa, whoa, whoa, I'm a content marketer.
[00:39:56] I'm interested in AI, but I don't know how to code. I don't know, machine learning. I don't, I'm not super technical. That's not what is in any of these chapters, these chapters are showing you what is possible today with AI and you do not have to go. Start some huge technology project. You do not have to go honestly, to start learn any new skills, except what's in this book of ways of thinking.
[00:40:22] And then you can, many of these tools are point and click tools. They're very straightforward to use. If you are diligent about picking the right ones up front, you can be using any of these tools for the most part today, uh, without any type of technical
[00:40:38] Paul Roetzer: back. We always get asked, like how much data do I need?
[00:40:41] And again, depending on the use case, maybe none like mm-hmm , if you're using a copy tool, like a copy.ai or Jasper or hyper right. Or go Charlie, or, you know, any of these tools we feature, um, You, you just, you just, the input is your website. Like, okay, I wanna write some social ads or I wanna create some social media posts or create an ad.
[00:41:00] Just you just give it the URL of the blog post and it generates outputs. So in that case, you don't need a bunch of CRM data. Now, if you're trying to do personalization at scale, of course, like you need much be more robust data. You need to connect data sources, but there's all kinds. I mean, dozens or hundreds of use cases.
[00:41:20] Or you're either using anonymized data from the vendor, or you're just creating content from something that's already there. Um, like a conversational agent. What's the phrase.io. I think we feature them. Yeah. You can have this knowledge assistant on your site and just learns from everything you've previously published.
[00:41:35] You just add the script to your site. It, you know, goes through the site, crawls, it learns from it and it could start answering questions like same day. So there's lots of AI tech. It doesn't have to be expensive and it doesn't need a bunch of data depending on what the use case. For sure. And then that leads us to the scaling AI chapter.
[00:41:54] So we chapter 15, 16, and 17 to wrap it and scaling AI. We won't spend a ton of time on today. It's a much, it's a bigger topic, but basically once you're done and you've proven success with a bunch of pilot use cases. Now, now the real work starts, and now you're starting to scale the business. And you're thinking about your data structure and how the different data feeds come in and work together.
[00:42:15] Um, how you train your team, how you get buy in from the C-suite, how you build a more human brand. Like we go through again, 10 step framework that you can use to scale AI. So it's a very important chapter. If you lead a team of marketers, if you are VP CMO level CDO, um, that chapter is for you. What I will say is it's very hard to find marketing leaders doing the things that are outlined in that chapter.
[00:42:44] It is a, um, In some ways, an aspirational chapter of what's possible for marketing leaders, but it's, it's challenging to find companies that are truly scaling AI today. Now that may change in the coming, you know, months and years, but don't wait for your peers. To publish the case study of how they scaled AI.
[00:43:08] That doesn't exist right now. So we are, as an industry, we are writing this chapter together in a way where we're actually going to be building the case studies of the companies doing this. Um, so yeah, like don't feel like you are behind when you get to the scaling. I's like, oh my gosh, this is so much tackle.
[00:43:25] I can't possibly do this. Just start at step one, just like stacking those success stories. Uh, and the other stuff can come. For
[00:43:35] Mike Kaput: sure.
[00:43:36] Paul Roetzer: Do you wanna
[00:43:36] Mike Kaput: maybe go ahead. Sorry, go. I was gonna say, do you wanna maybe dive in from there to the more human
[00:43:41] Paul Roetzer: chapter? Yeah. So that's the, the one that, you know, the last two is funny.
[00:43:46] Like my first book I wrote, um, pursue purpose was the last chapter and it was a chapter. I almost didn't write cuz it was very personal to me. Um, like why I built an agency and what I thought was possible with agencies. And I kind of did this similar thing with this book. Like the last two chapters are very personal to me.
[00:44:02] They're um, they go to the root of like why I wanted to build what we're building and why I think it matters. And so the more human one in particular, you know, we, in 2019, when we launched the marketing AI conference, the tagline was more intelligent, more human. And the premise there was, if we succeeded at moving the industry forward and teaching the industry about AI and ways to do marketing smarter, We could also open up the possibility of people hacking marketing and taking a bunch of shortcuts and manipulating human emotions and behavior with AI because it's all doable.
[00:44:38] And so the more human side is basically saying as an industry, we need to accept the fact that AI can be used. For negative outcomes, it can be used by bad actors to do bad things, but the more marketers understand what it is and what it's possible, the more we can work together to create ethical standards for the use of AI, and we can focus on AI for good.
[00:45:01] And so we talk about, I think we use the example of Adobe, uh, their ethical AI for ethics, um, standard policy. We're at marketing AI conference. This year. We have a responsible AI workshop in August. Where we're going to actually create, you will come out of it with an ethics policy for your company. Um, we don't know of one that exists in the industry specifically for marketers, so we're gonna, you know, create it.
[00:45:23] Um, and then we end with, there's probably like 20 different organizations that you can follow. If this is an area you're interested in, um, That are all working on this exact problem that, uh, people know people who spend their life in AI research are aware very acutely of the downsides, the dark side of AI.
[00:45:43] And there's some very smart people in very powerful positions who are trying to leverage their influence to make sure that AI is used in a positive way. So I think some people, when they get to this chapter, They will already understand this could go wrong, that there are certainly ways AI can be used negatively, and we highlight a number of those throughout the book.
[00:46:05] And so I think it's very important to us that more people are asking the hard questions about AI, like related to bias and ethics and more moral standards, um, and working towards a positive outcome for every.
[00:46:20] Mike Kaput: Yeah, that's awesome. I mean, I couldn't obviously agree more and just as a kind of final point there, I mean, we've seen.
[00:46:27] So many, unfortunately real world use cases of what you just described. I mean, how many people, that's probably a very accessible way for marketers to be seeing some things around AI, as everyone has heard of what has happened, perhaps at a Facebook or at certain areas or people using certain technologies to do bad things.
[00:46:48] So it couldn't be more important to be thinking about it now, before it's
[00:46:52] Paul Roetzer: too late. Yeah. And I think my general guidance people is. Dwell on the negative stuff right away. Like you really need the broader understanding of AI before you can. Really get caught up in the dark side because the dark side's there and it's not going anywhere.
[00:47:07] It'll be there for you when you're further down your learning journey. Trust me. Like once you learn, what's possible, you can't unlearn it. And, um, so I think it's just really important that people understand the fundamental elements of AI and what it's capable of doing. And then when you're ready that you start asking the hard questions of your own company, of your industry, of your own practices, because.
[00:47:30] It can be used to, to do some very powerful things. Um, and so I just, I really, I want more people thinking about this and I think that's the goal of that chapter. That's awesome. And that takes us to the last one, the AI and new chapter. And that was, that was probably the, my favorite one to write. I actually wrote that chapter in about 30 minutes, I think.
[00:47:53] Hm. Because it was just one of those where you'd written everything else. And I was trying to just. Think about the different people who would be reading it and try and tell the story of like, okay, you're a college student, you're a marketing professional. You are a marketing leader. You are a CEO, like, and.
[00:48:13] And put myself in their shoes of where are they right now? What are they feeling? And thinking after having absorbed all this information and what do they do with it, for their, for their own good, for the good of their company. And that's really where that chapter comes from is of, you know, for me, a very personal.
[00:48:31] Here's what I think I didn't do research for that chapter. I didn't go like read a bunch of articles and try and curate it. That was just me writing sort of from the heart of what I thought was going on and what I thought the opportunities were. So I, I mean, as a writer, Mike, you can appreciate like that.
[00:48:46] That's, that's when writing is it's most fun when you're just editorializing on things you think about and believe in, um, and you hope other people care. .
[00:48:55] Mike Kaput: Yeah, absolutely. I couldn't think of a better way to end this book. And I love the idea that as a writer, you know, when it just flows naturally, that's some of the most important things you might be saying, because no matter what, you have to be thinking about these issues to write them clearly.
[00:49:12] So if you did that clearly, that has been on your mind for a long time.
[00:49:15] Paul Roetzer: Yeah. Yeah. Yeah. And I don't know that ever, you know, it's, it's always hard when you've written so much and you've given so many talk. To, to say things differently. And so I've definitely found sometimes you need to forget everything else you've previously said and, and just clear this slate and don't go look at the last article or the last podcast transcript or anything like that, and just like free your mind and just, just write it.
[00:49:37] And I, I know as a writer, I historically have gotten caught up in that. Well, let me go reread. To 10 things, then I'm copying and pasting paragraph here, paragraph there mashing things together, rewriting and all of a sudden it's like five hours past. And I haven't written anything new versus just like clear the slate, go to the blank page and just write what you're feeling at the time.
[00:49:58] And both, both approaches work. But I think for a chapter like this, it just needed to be something new and yeah, for sure. Um, and then there's a conclusion that actually, I, I, a lot of additional thoughts came out in the conclusion was funny. I thought I'd I was done. I thought the whole story was out of me.
[00:50:16] And then I, I got to the conclusion and I. Kind of a similar thing. It just like started writing and I was like, oh, okay. Like, I guess there's something else I needed to say. So the conclusion's, uh, hopefully valuable for people. And then we, you know, kind of wrap it with 12 things to know about AI. And I think you could probably just take that page as a starting point.
[00:50:36] And then one, the thing I'll advise people is in the conclusion, like I kind of wrap it with like a 500 word editorial and then it just goes chapter by chapter with like five to seven bullet points from each chapter. Like you, you could theoretically like skip ahead and. Read those summaries and it would give you a really good sense of what's in each of those chapters.
[00:50:56] Um, so yeah, I mean, again, it, it's, it's a decade of my life. Mike's been working on this stuff and writing about it with me for probably since 2014, um, back, you know, our agency days together. So he and I have both put, you know, the better part of a decade of our lives into trying to understand this and trying to find ways to make it make sense to other people.
[00:51:17] So I, I hope that comes through in the book. I. Um, you know, again, we use it to advance your own career, use it to build a smarter business. Like that's what we want. We want to hear those stories. So, you know, reach out to Mike and I, you can get the book@marketingaibook.com. We'll take you to the main page on the site.
[00:51:37] You can get it at Amazon. Hopefully you can find it at bookstores. Um, as a author, you never, you have no control over that. unless you pay for placement at the. Bookstores. Uh, it's a whole nother story. Um, but yeah, you can get the audio digital and print on Amazon on apple. Um, heck buy, buy all of 'em if you want.
[00:51:58] that's what I do with good books. I usually get the audio book first. It's like, that was really good. I need a hard copy and then I want the digital copy so I can highlight it. So I usually buy all three if I really like a book. So you get a special reward from us. If you buy, I don't know what it is, but like email me and tell me, and I will send you some reward.
[00:52:16] Uh, any closing thoughts on your end, Mike?
[00:52:20] Mike Kaput: No, I mean, I'm just extremely excited for June 28th. I mean, it's cool to have a book launch coming up and I truly, you know, I know I'm self-promoting here, but I truly think if you're a marketer and you are interested in this stuff, there couldn't be a better book to take you from point a to point Z.
[00:52:36] And I just highly encourage you to give it a shot.
[00:52:40] Paul Roetzer: Yep. So, yeah. Thanks for listening. Hopefully this was interesting from a few levels, like the book writing process to me, if, if you've ever done it, that trust me the first time I had to do it, you learn all kinds of stuff. And it's very different now than it was in 2011 when I wrote the first one.
[00:52:55] So yeah, I mean, hopefully that's interesting. And in some of these insights, even if you don't end up buying the book, you know, hopefully you learn something along the way today and, uh, yeah, it's not too late to join us for the marketing AI conference as well. You know, you just, if, if this is hitting at the right time for you and this stuff's all clicking and making sense, and you're seeing the opportunity to build a smarter business and, and grow your career, then, then visit macon.ai.
[00:53:18] It's justi co n.ai. And, uh, there's still time to join us August 3rd to the fifth in Cleveland, Mike and I will be there and a few hundred of our, our friends who. You know, NextGen marketers as well. So thanks again, Mike, thanks for, you know, obviously your amazing contributions to the book and everything we're doing here.
[00:53:37] And, um, and to Matt Holt again, and Ben Bella books, who've been an awesome partner in, in this process. And, uh, we'll talk to you next time. Thanks so much for being a part of it.
[00:53:49] Thanks for listening to The Marketing AI Show. If you like what you heard, you can subscribe on your favorite podcast app. And if you're ready to continue your learning head over to marketingaiinstitute.com. Be sure to subscribe to our weekly newsletter. Check out our free monthly webinars and explore dozens of online courses and professional certifications until next time, stay curious and explore AI.