Episode 174: How Marketing AI Boosts Your Competitive Edge

You run an agency.

Yeah.

So you are the quintessential entrepreneur and you're acquainted with this topic from an owner's perspective, not just a marketer's perspective or from the AI Institute, but what are you seeing in B2B?

Yeah, I mean, everybody's just moving too slow. Like, I don't, I don't think that overall people comprehend that this isn't like sci-fi. It's not future looking. The tech is real. Like, we don't need leaps forward in the tech to dramatically improve the performance of marketing.

Welcome to Skucast, the podcast for entrepreneurs in the promotional products industry. Skucast shines a light on our industry's best work, features maverick personalities, and discusses what's really involved in running a modern promotional products business. SkuCast is the official podcast of Common Skew.

In 2021 alone, Gartner projects artificial intelligence will create $2.9 trillion worth of business value and create 6.2 billion hours of worker productivity globally. Last week, Drift and the Marketing Artificial Intelligence Institute released the 2021 State of Marketing AI Report. It's a massive research project that included insights from marketing professionals on how AI is impacting their businesses and driving results for the future. My guest today is Paul Raetzer. Paul's the creator of the Marketing AI Institute and is the founder of the agency PR 2020. He's the author of a book many of you know, The Marketing Agency Blueprint, and the book, The Marketing Performance Blueprint. Today I talk with Paul about the results from this landmark study, but I chat with Paul not just as the preeminent leader of AI progress in marketing, but also as an entrepreneur and marketing professional himself. We talk about using AI to drive revenue, to radically accelerate your team's productivity, why AI is crucial for strategic thinking and planning, and how to implement AI in your business in a way that transforms big theory into practice. And since Paul authored the book on the Marketing Agency Blueprint, while he's with us today, I talk with Paul about the agency model and particularly the pricing conundrum around charging for ideas and charging not just for our work, but charging for our value. Hi friends, I'm Bobby LeHew, the Chief Content Officer at Commscue. Before we get into our chat with Paul, I wanted to let you know that registration for Commscue University Plus is open. TSU+ is a 2.5-hour event specifically for CommsQ users who are serious about becoming pros on the platform. With 3 focused sessions plus breakout electives for sales reps and owners, your team will learn how to maximize productivity while gaining insight into how other pros use CommsQ to grow their sales. The strongest benefit of attending CommsQ University+ is that you get a masterclass level of learning concentrated into just a few hours. Built for teams, it's a minimum investment and maximum impact. You can check it out and register at commonscue.com/universityplus. This episode is brought to you by CommonsCue, the work-from-anywhere platform that powers your connected workflow, enabling you to process more orders and dramatically grow your sales. To learn more or to begin your free trial now, visit commonscue.com. Now here's my conversation with my friend Paul Raetzer. Paul, two statements I'd like to ask you to comment on because I think they're powerful and they cause us to lean forward a little bit. One is the future is marketer plus machine. Can you explain that?

Yeah, I mean, the way we look at artificial intelligence, this ability to intelligently automate human-powered tasks. So when you think about your job and if it's in marketing or sales or just in business in general, there's so many things you do that are repetitive that we are manually trying to do as humans. That's where machines excel is doing repeatable processes and then getting smarter with how to do them better as they're going. And so I just look at so many things that we do every day where it makes so much sense for an intelligent machine, you know, this idea of AI to enable us to be more efficient at our jobs and be better at them.

There seems to be this gap in people's minds where we think of AI as reductionary, as reducing it down to just the most banal tasks. And there's this phrase that— and when you look at the State of the Industry report that you just released, the tasks that AI is capable of doing now is mind-blowing. There's a phrase that says we have entered the age of intelligent automation. Can you explain the difference between automation and intelligent automation.

Definitely. So if you think about the last, I mean, really everything you've done in your career up until now, whether it's email marketing or social media or CRM or automation, like setting up workflows, you write all the rules. Like nothing in marketing happens without you telling the software what to do.

Right.

And so again, whether you're figuring out when to schedule the social shares or what subject line to use in the email or what color to make the call to action button or how to edit the video, like all of those things you had to learn how to do, and then you only improve on them if you go learn more. The software itself doesn't get any better. And so intelligent automation is the idea of automating a task, meaning setting, normally it's if-then statements, if this happens, do this, or telling it what to do, but having AI involved to where it actually can improve on its own and learn how to do that task better. So intelligent automation is just AI plus automation. But historically, marketing has been all human all the time. Everything has been human-powered. All the rules of how things work have been written by humans.

Is this the same difference between machine learning and deep learning?

No. So the best way to think about it is artificial intelligence is— the best definition I've seen is the science of making machines smart. Meaning machines, hardware, software, doesn't know how to do anything without being told what to do. So we've adapted that definition to marketing AI as being the science of making marketing smart. But AI is just this umbrella term for tools and technologies that make the machine smart. The primary subset of AI is machine learning. And what machine learning means is it takes data in. Again, it could be any kind of data. It could be in your CRM system, your analytics, and then it makes predictions about outcomes. So it's the machine learning and then using this data to predict what's going to happen. What's going to get someone to open an email? What's going to get them to click on a social share? What time are they going to open emails? When should I send this email? When should I recommend this piece of content? It's using machine learning, taking in data to predict a behavior or an outcome. Deep learning is a subset of that that's actually trying to give the machine human-like abilities. So if you think about your iPhone, how it can be unlocked with your face, that uses facial recognition. So it's giving the machine the ability to quote unquote see. When you talk to Siri or Alexa, it's giving the machine the ability to speak and to understand language. So sight, language generation, movement, those are all things that machines don't have. Deep learning is the way we do that, and that's by actually trying to teach a machine to think think like a human brain.

As we talked, this audience comprised of marketing professionals in a very broad sense. Some of them are entrepreneurs running businesses, large businesses, small businesses. Some, many are marketers. They're basically all running marketing agencies though. Yeah, of some point. The 2021 report stated that most marketers were in one of three phases when it comes to actually implementing AI. So they're in researching, understanding, or piloting. And I think we can grasp the researching and understanding phase, except for I do think there's a lot of misnomers about AI that we're dealing with now. But what is the piloting phase and what is it that is most important for us to grasp?

Yeah, so what we always talk about is to get started with AI, you need a pilot project. Like, you need one thing to test to see if it works. And so the idea of a pilot project is to take a distinct task or project and use AI to try and do that task more intelligently, more efficiently, achieve greater outcomes with it, and then to prove over time that you can stack these tasks on top of each other and start using AI in a bunch of different ways. But for most people, because AI is— it's programmed to do a very specific thing. You don't go get an AI platform, flip a switch, and now AI just does everything for you. It's those examples I said, like if you're going to send an email newsletter, you might buy a product just to help you write the subject lines that, you know, it'll recommend subject lines to write and tell you which has the greatest probability of driving open rates. Or for us, we use one that does advertising creative. We upload 3 different versions of a potential ad. The machine runs models of that ad and tries to predict which one will perform best based on the text and the images and the colors and all the other factors it considers. Without us spending a dime on actually A/B testing anything. Yeah. It just runs these theoretical models. So those are pilot projects. That's like, okay, we do a lot of advertising. We spend a lot of money on Facebook. Let's find a smarter way to do advertising on Facebook. And so that's piloting is like you're actively looking for a couple of projects that you can try AI with.

So taking it out of theory and putting it into practice.

Exactly.

The two most common outcomes from those that are proactively piloting AI solutions in their business, and I think this is gonna get most most of our audience's attention at this point. Yeah, but from the study, those that responded said the two most common outcomes were number one, accelerating revenue, and then number two, getting more actionable insights from marketing data. You shared the ad. Can you share another example or an outcome from each of those?

Yeah, so like let's say we use WordStream as an example, also in the advertising space. But I mean, a lot of these early projects are, are going to be advertising for people because that's where a lot of marketing dollars goes. But what WordStream does is it looks at the campaigns we're running on, say, Google, and it actually finds anomalies and insights within campaign performance, surfaces those insights, and actually recommends to our team how to improve the performance of the campaigns. And so rather than our team having to learn how to analyze the data coming out of the ad campaigns, the machine actually does the heavy lifting, and then the human looks at those recommendations like, yeah, that makes sense, or that doesn't make sense. The machine's not taking the action necessarily as a result. It's just telling us here's what it thinks we should do. So that would be another one. And then we also use a, a tool called Market Muse on content strategy. And what that one's doing is looking at all the content we've previously published and trying to find posts that if we enhance them with some specific answers to specific questions, have a greater probability of driving more traffic than other posts on our site. So it helps us prioritize what to enrich and republish on the site.

I will be looking at that one.

Yeah, it's awesome.

Yeah. I'm curious too about turning— this is what you mean by turning data into intelligence too.

Yeah. Yeah.

What are the 5 Ps of marketing AI?

So what we were doing originally is we were trying to find ways— a helpful way to understand AI is to look at just a bunch of use cases. And what I mean by that is a bunch of tactics, a bunch of things that you do as a marketer. And so we originally thought about categorizing them like advertising, content marketing, email marketing data, and we do that still, but we were trying to find like a higher-arching framework for it. And so what I ended up doing was settling on this idea of 5 Ps. So there's planning, production, personalization, promotion, and performance. And so underneath those, you can start to think about What you do for strategy, so planning. So like, for example, pricing strategy might figure under there, you know, figuring out budget allocation. Those are planning things. Production is like the creation of intelligent content. Personalization is being able to personalize experiences and content at scale. Promotion is obviously, you know, running the promotions that actually drive the actions. And then that last one you mentioned, the turning data into intelligence, is the performance side. We all have data. We all have analytics coming from every tool we use. How do we actually take that data and make it actionable? That's what we mean by like taking all this pipeline of information coming at us in these data points and charts. It's like, what do I do with this? What does this mean? Like, what are the actions I can take?

Taking just one of those piece, the planning part, how do you envision AI reducing the amount of time spent on work about the work or all that time spent planning and managing the work being done?

Well, when you think about planning, I mean, in general, what you're doing is you're trying to predict what actions will drive the outcomes you desire. And whether you're doing budget allocation or where do I spend my marketing dollars, or, you know, you're trying to figure out how many pieces of content you need to create and what channels to promote them in, it's all coming down to predictions. And so a lot of marketers, you know, use instinct and experience to guide their strategies. Right. And sometimes they use data. You know, you'll look back at what's worked, what hasn't. My belief is that strategy largely comes down to trying to predict outcomes based on past inputs or data. And so I just feel like everything, whether it's your pricing strategy or your channel distribution strategy or whatever it may be, that you should— machine learning, the part that takes data in and makes predictions about outcomes, It seems like it should be infused into every part of marketing strategy, but it's not today.

In the state of the industry report, there were a list of top 10 use cases, and a few that got my attention were predicting customer churn and recommending highly targeted content to users in real time. But where do you think we have the most gains to make in terms of application that can make the most strides for us as marketers? So I think we have an audience listening that that may be thinking this way. I don't be too presumptive, but I think they may be thinking, I get it, I see how powerful AI is, but I'm— there seems to be a chasm between that big topic of AI and then how I'm going to implement this in my business to give me an edge in 2021.

That's why we talk so much about the use cases, and in the report there's 49 different use cases that we had people evaluate the value of to them to be able to intelligently automate a task. And I think the point we made in the report is it's so subjective. Like, the value you can get from AI largely comes down to what do you do every day that you have the ability to do more efficiently. And so there's great studies though, like the one I mentioned in the report is the McKinsey Global Institute, and I think it was from 2018, but they actually looked specifically at marketing and sales and said, what is the potential of AI? And then they lumped in other analytics, which is advanced forms of machine learning, basically. And they tried to say, what is the market potential for it? And they came up with $5.9, rounded up to $6 trillion in annual impact, just in 7 categories. And so the dominant one for them was pricing and promotion at $1.9 trillion. And this is annual impact. And the idea there is like, a tangible way to think about pricing and promotion is if you think about airline pricing, And if you and I both go on Travelocity right now and search for the same flight, my guess is we're going to actually see two different rates because what they're doing with their pricing is they're doing real-time dynamic pricing based on past behavior. So they know if you or I have looked at that trip before, they know how close we are to the trip, they know the capacity of the plane. They have all these variables that go into what price they're going to set at that exact moment for that exact flight. Flight. Now, we can't do that. Like, humans can't process that kind of data. And no matter what business, even in promotional products, like, should we just set a price and forget it? Like, does weather affect what people are willing to pay, right? There's all these other variables. And so if you think about industries that you can move the needle on your profit one way or the other based just on your pricing strategy, and if you can optimize that pricing strategy like, again, fitness centers, hotels, like whatever it is, there is massive amounts of value to be unlocked because the people are going to be willing to pay it because it's an optimal price point for that person at that moment. So that's one example. The next product offer recommendation engines are huge. But yeah, I mean, budget allocation, churn reduction are pretty much universal.

What are you seeing with B2B in terms of use cases, and what has sort of opened your eyes with B2B? And we'll get into these, a few subset of questions here, because What many may not realize is that you run an agency, so you are the quintessential entrepreneur and you're acquainted with this topic from an owner's perspective, not just a marketer's perspective or from the AI Institute. But what are you seeing in B2B?

Yeah, I mean, everybody's just moving too slow. Like, I don't, I don't think that overall people comprehend that this isn't like sci-fi. It's not future looking. The tech is real. Like, we don't need leaps forward in the tech to dramatically improve the performance of marketing. So, you know, there's good examples, but they're hard to find. The B2B companies that are doing it don't want to talk about it because it's a competitive advantage right now that they've done. But what the best thing you can do is go look at the big players in the industry and see who they're doing it for. Like, I know we just did our webinar with Drift and they talked about Okta was an example of a company that was using the AI capabilities within Drift to dramatically accelerate revenue. Uh, Persado is an interesting company in the language generation space, and you can go look at some of the big players that they're working with. But that's really the best way to do it. But honestly, we struggle to find case studies, especially in the B2B space of brands that are doing it and are willing to talk about it.

Yeah, we will link to everything that Paul has mentioned in this conversation, including the link to the study. Uh, I'm curious, Paul, what insights did you gain working on this report. I mean, what surprised you that you weren't expecting as you were working on this?

Yeah, so one thing that, you know, I've done over 100 talks on AI in the last 5 years, and universally people are scared. Like, you know, you do this kind of macro level, what is it? You show a bunch of use cases. People sit in the crowd, they see a use case up there that they spend half of their time doing, and it's like, oh my gosh, like, am I going to have a job in 5 years? Like, I get that question all the time. So I was a little surprised that only, I think it was 15% of people identified fear of AI as a barrier to adoption. I would have thought that would have been higher.

Do you think that's an evolution just recently?

I think that it's just people are starting to understand that it's not sci-fi, that it's just math. Like, it's just, you know, it's just making predictions and it's doing repetitive tasks that we don't like doing anyway. Like, I hope that that's what it is, that people are starting to realize this isn't going to just shake everything up and everybody's going to lose their jobs. It's actually in a lot of ways additive. You know, it complements what we're capable of doing as humans. Yeah. In that same question, though, you know, we, we offered like 15 different options as to what the barriers are. And the dominant response by far, by 30 percentage points, was lack of education and training. And I guess I didn't know, like, I didn't think the industry was far enough along to recognize that as the obstacle. I thought we were still like, yeah, we just don't have enough awareness or we don't have executive support for it. There's a lack of vision in our company. I expected more general responses like that than saying, like, yeah, we just don't— no one knows what to do. Like, we're not sure where to go. I was a little surprised on the piloting at 34%, that 34% said they were at that stage. I would have guessed that was lower. And then the other one I think is just the beginner and intermediate. So we had 50% said beginner, and I think it was like 37% said intermediate. I'm not sure people are being honest with themselves. And I don't mean that in a negative way. I think if 37% are intermediate, I would love to talk to those people because I don't, I don't know what they're defining as intermediate. My perception is it's like 95% beginner. Like, we're just not there as an industry. I think there are a lot of people starting to research it, trying to understand it. I believe those data points very firmly, but I feel like we're so early, and that's exciting to me.

As I mentioned earlier, you run an agency, and I don't quite know how to put this question, but here's what I want to try and ask. How do we engage our teams in a way that gets them to ask, how can I do this smarter? Is this a part of the challenge? You mentioned, for example, earlier that we do a lot of our planning and strategy by instinct or by history. And we're not asking almost like a constant question of how can I do this smarter, more efficiently, faster?

Yeah, it's tricky. You have to teach the mindset. So what I've found, because I started working on this stuff in 2011, and I would say it was 2015 or '16 before I would have been confident to stand on stage in front of a group of machine learning engineers and say what I thought machine learning was. Like, it took me a long time and a lot of reading, a lot of conversations with AI experts to realize that most of them were actually struggling to explain what it was they do and what the technology does. So once you understand it though, and you actually simplify it down to this whole idea of predicting outcomes and, you know, human-like abilities of language and vision, it's not that hard to comprehend what it is. It's really not that abstract once you, You understand a base level. But once you do, you look at everything totally different. Like, I look at inefficiencies in our agency, and it just drives me crazy because I see things like, why are we doing this? Like, why did we spend 5 hours doing this? Like, I know for a fact we could use AI to do this. And so the way we're doing within our agency is we actually rotate training intensives. So we're a huge HubSpot partner, like, in terms of our focus, and we drive growth for clients through HubSpot. And then AI is sort of the future for us. And so what we do is we rotate 3-month intensives. So for 3 months, the team is focused on HubSpot training and certifications, and then the next 3 months, they're focused on AI training. And we create that curriculum through a mix of our own content and training and then external courses and information. And so we create this curriculum to try and give that base level competency to our team so that hopefully they look at things differently.

How would you advise a business owner, an entrepreneur, a marketing professional to activate AI in their life and in their business?

Yeah, you just need to keep taking the next step in the journey. I understand that it's abstract. Again, it took me years to be able to make it approachable as a topic. Lots of writing, lots of research, lots of speaking, honing the message to see what makes sense. So I get that it's abstract and it can feel overwhelming, But I promise that if you just take the next online course or start with the State of Marketing AI report as a great entry level for people, and then just like if it sparks an interest in you and you start to see the opportunity, then go read the next thing. And like, don't try and learn it all in 30 days. It's not going to happen. But commit yourself to a regimen of like gradually improving your knowledge and capabilities in that space, and you'll be rewarded.

Yeah. Switching gears just a little bit, you and I met several years ago when you wrote Marketing Agency Blueprint. We have some fans of your book listening today. And can I ask, why did you write it?

I thought it would help people. So, you know, people who know the story know we were HubSpot's first partner back in 2007, early 2008, and became kind of the origin of their partner program, which today has, you know, thousands of agencies around the world. So for a couple of years, in 2008 to 2010, you know, Pete Caputo is a good friend of mine, is now the CEO of DataBox. He was at HubSpot and he kept pushing me to share what we were doing. He's like, you know, I think, I think there's an ecosystem we've built here. I think HubSpot could build a partner program around what you're doing and other firms like yours. You should share it. And I was like, why would I share it? Like, why am I going to tell a bunch of other people what we're trying to figure out? Still, I haven't solved anything, per se. And he's like, all right, well, I think you should. So then December of 2010, I got a, I woke up to a direct message on Twitter from a lady in Romania, and she said she was building her country's first inbound marketing agency because of us. And I was like, that's kind of interesting. Like my wife's family's from Romania. So on the car ride into work that day, I literally decided to change directions. So I got to work, I emailed Pete, I said, all right, man, I'm ready to share. I think we can help a lot of people, a lot of entrepreneurs who are trying to figure this out. I'm ready. As long as HubSpot will support it, I'll do what we need to do. So fast forward 3 months and I was at South by Southwest 2011 and a good friend of mine, David Meerman Scott, people might know from the Marketing PR and a bunch of other books. He invites me to the Wiley party that night, the publisher. So I go to the Wiley party with David. He introduces me to his editor and A week later, I, I had a book deal. Like, so the basic premise was David explained to Wiley, he's like, no one has written the book on how to build a modern agency. That's what he's doing. Like, we're not there yet. We haven't proven all these theories and everything, but like, we've learned enough and we're close enough to this that I think we can help people. And so that's really what the book became is like, what is it gonna take to build a modern marketing agency? What's happening with HubSpot? What's happening in the ecosystem? And what are the opportunities for people to build a different kind of agency.

And that was really the— there was a shift happening at that time, right, in the market in terms of pricing for services and strategy and things like that. And for those that— folks that haven't read the book or even seen it, the premise was you were sort of shaking things up a little bit too, right?

Yeah, I mean, the chapter one is Eliminate Bill Blowers. So I started the agency in 2005 on the idea of standardized services and set pricing. I felt that billable hours were inherently agency-centric, meaning the agency gets rewarded for inefficiency. The longer they take to do something, the more money they make. Right, right. Usually— not saying that, that agency professionals are dishonest, but usually when there's an incentive to be inefficient, the larger group will likely take advantage of that incentive. So my feeling was we had an entire industry, entire financial models based on what appeared to be an inefficient pricing system. And so my belief was, if we can be more transparent with pricing, if we can say, here are all the things we do, here are what they cost, if you would like them, you know, we've agreed on a value exchange, you've already seen our price, there's nothing to hide from you. And so this is before marketplaces and everything, you know, that's transpired in the last 15 years. And so that was what I started PR 2020 based on. And so we eventually evolved and created what we call point pricing model. That's like a value-based model. I hadn't created that yet when I I wrote the book. But that was the basic premise, is agency services should be value-based. It's in the better interest of the agency and the client if there's an agreed-upon value, and then the agency's not leaving money on the table too. Like, my— a great example for me is crisis communications plans. Like, why should you pay by the hour if I solve a crisis for you in 5 hours, right? Why should I only bill you 5 hours when it was a lifetime of work that let me just save you a billion dollars in market cap?

Like, right?

So what is the value of an hour in that scenario? It doesn't hold up. So it actually goes both ways.

I have the most basic question then. How do you charge for ideas in that sense? It's almost as if you have two minds. One mind is, here are my operating costs, here's the money I want to make, and I'm going to charge by the hour. And the other mind is— and, and that's almost solely has to do with raw costs of service. The other is the potential, like what it gains for the brand or what it gains for the company. And I know that's kind of what you were just talking about, but how do you charge for ideas?

Yeah, it's a mix. So the vast majority of what we do is, is has a point value to it, and a point's just a fixed unit of value. So, you know, it could be 5 points, it could be 13 points, it could be, you know, 21 points, whatever. And that enables us to create what the cost of a project is. Now, we consider hours in there. Like, we would look and say, okay, well, realistically, this is like 5 hours. It's gonna be a mix of senior talent and, you know, junior-level talent. So we think about the hour investment. But then you look and you say, is there a value accelerator to this though? Is that client benefiting from the fact that we've done that project 5 times in the last year for other clients? And even though we can be really efficient with it, the value to them is still far greater than what we would charge by the hour. So that's where we look at it and say, okay, there's just a higher value here. But by having the point model, we can actually project out over a year and distribute those points and say, here's how it should kind of break out depending on what we're doing for them. So we definitely factor in what is the end outcome value to a client for the service that they're asking for.

Is that subjective or is that sort of a price you work on in tandem with the client? So just going back to your example, of crisis management plan, the value for that company is tremendous.

Yeah, it's subjective. I mean, we try and standardize the points as much as we can, but another good example, like a really tangible one, is a blog post. So if we write a blog post and this is one I haven't honestly cracked the code for, like how you do this other than just having different pricing. But if I wrote a blog post for a client that has 1,000 visitors a month to their website and almost no social presence, that blog post isn't worth anything. Like, it's not going to do anything for them. It could be the best blog post ever written in their industry and they have no way to amplify it. There's nobody there to read it. Right. However, take the same blog post in the same industry and you write it for someone that gets 100,000 visitors a month to their site and has 40,000 subscribers to their newsletter. Yeah. You put that blog post out, gets in front of 40,000 people, 30% open it. 12,000 people are now amplifying, or at least aware of absorbing that information. I would argue that the value of that blog post is greater based on the existing platform the client has. Yeah. So you could see a scenario where you could actually charge differently based on what you're able to do as an outcome for that company. Yeah.

Outcome— I love the outcome-based thinking about that. Does your agency then charge on this point system per se? Gone are the retainers of the old world, or you still do a mix of both?

So what most of our clients are on is a monthly recurring revenue model, like a contract model where they buy an allocation of points each month, and then we help them figure out how best to distribute it. So if it's $8,000 a month, I don't, I don't know exactly how many points that is, 57 points or something like that. They get that. Now we will build in their plan, here's how we recommend based on your goals and, you know, your existing platform and everything you're doing, here's how we would recommend allocating those points.. And that they, you know, they can obviously provide feedback and guidance, but then we just allocate those points and then they're each fixed to a project. And then the client knows, okay, I've allocated my 57 points and out of that I'm getting these 17 things this month.

What has changed in the market and agency since you wrote that book? So if you had to write it again today, and we'll put a link in the notes again for folks to find the book, if you had to write it again today, What would you emphasize due to these changes in the market?

So I wrote the manuscript for the book April 15th to July 15th. I turned in the manuscript in 90 days. 2 months prior to starting that manuscript was when IBM Watson won on Jeopardy. That was the moment that I turned my attention to artificial intelligence, that I started wondering what that technology was and what was going to happen. When it came to marketing. It was not in marketing yet. It was a couple years before the major players started even hiring machine learning engineers. So we were in that all human-powered automation phase of marketing. Marketing automation was blowing up. Uh, HubSpot IPO'd in '14. ExactTarget got acquired for $2.5 billion by Salesforce. Like, billions of dollars are moving around and nobody's doing AI. So what changed was Everything changed. Like, what I thought was possible changed in the industry. So the book is still highly relevant. It's still a bit evergreen in terms of the principles of building an agency. The thing that has evolved is I think agencies can become smarter. I, I think AI technology is gonna transform the way they operate, the efficiency with which they operate.

It'll change the skill sets needed, and it'll change the service Yeah, it's interesting the way you're wired, Paul. This is just an observation. I don't know if there's a question here, but you are hardwired to look for improvement, operation efficiency. So you do that with your agency. You saw AI and you gravitated toward that.

What's next for Paul Reitzer? Uh, that's a really good question. So I, I think when I set out to build the agency, I wanted to build some— something significant that could help a lot of businesses. My, my original model was actually— my wife was an artist and my parents owned a franchise, Cookies by Design franchise. And so I saw the need for the mass market of small businesses to have help that I didn't think agencies were structured to help them. And so I thought, well, if I can build a more efficient model and I can borrow from the manufacturing ideas of achieving economies of scale in the creation and distribution of services, then I could actually help more businesses. Quickly realized how hard that is to do as an agency, that it's really hard to build a service business that can service small businesses at scale. So years go by, I get, you know, caught up in the AI thing. You know, I think the AI thing is just at the beginning. You know, when I look at where we go from here, I, I think we're just at the very, very leading edge of this next frontier in digital transformation. Powered by AI. I don't know exactly what we'll do with it, but the Marketing AI Institute we've created is really a media company, an event company, and an online education company rolled into one. But our whole mission there is to make AI approachable and actionable, to help agencies, to help businesses, to help marketers figure out how to actually understand and apply AI. And I think there's a long roadmap ahead. To achieve what we're setting out to do there. I won't stop probably looking at those inefficiencies, but I think AI is going to help solve a lot of them. And in a weird way, I think what I tried to do 15 years ago to help all these businesses, AI may have been the missing piece to the puzzle. I don't know how yet, but a lot of things are going to be possible that weren't possible 15 years ago.

That's awesome. Well, Paul, it's been a privilege knowing you through the years, and it's fun to watch your brain work. And thanks for joining us on the SkewCast today.

I loved it.

Glad to do it anytime. Thanks so much for tuning in to this episode of SkewCast. Be sure to keep up with our latest content by subscribing to SkewCast on iTunes or to our blog at community.comonskew.com. Until next time, friends, thanks so much for listening.