<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=2006193252832260&amp;ev=PageView&amp;noscript=1">

35 Min Read

[The Marketing AI Show: Episode 1] Salesforce Leader Discusses the Post-AI Consumer

Featured Image

Wondering how to get started with AI? Take our on-demand Piloting AI for Marketers Series.

Learn More

The Marketing AI Show—the podcast that helps businesses grow smarter by making artificial intelligence approachable and actionable—has officially dropped!

You can now listen to the first four episodes of the podcast on your favorite podcast app. Keep reading for more on what to expect in episode one.


Episode 1: Mathew Sweezey, Director of Market Strategy, Salesforce 

Sweezey, MathewMathew is regarded as one of the leading minds on the future of marketing and his visionary insights into consumer behavior, technology and new business strategies have changed the way startups, Fortune 500 and nonprofit organizations alike find customers, and build modern brands. 

In addition to his work with brands, Mathew is the host of the award-winning podcast The Electronic Propaganda Society and an accomplished writer having written for The Economist, Forbes, HBR The Observer, and Adage. He’s also the author of The Context Marketing Revolution

During his interview with show host Paul Roetzer, Mathew discusses the “Post AI Consumer,” and how AI has already changed the way consumers connect, find things, interface with technology, and make purchasing decisions.

Watch the whole interview to hear more about:

  • How “deep fakes” can actually be a positive thing. 
  • Shopping on the "edge," and why it’s the next major wave of commerce.
  • How you can adapt to a world in which AI powers everything.
  • Ways to understand and apply AI in your career now.


[Video] Watch the Full Interview 

Read the Transcript

Disclaimer: This transcription was written by AI 🤖, thanks to Descript

[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. Thanks for joining us. This is episode one of the Marketing AI Show. We're kicking off the podcast with Mathew Sweezey, Director of Market Strategy at Salesforce. Mathew states that by 2025, marketers will become AI reliant.

[00:00:42] Meaning AI will transform creativity, strategy and execution. He also talks about the post AI consumer and how AI has already changed the way consumers connect, find things, interface with technology and make purchasing decisions. So what can you do to understand and apply AI in your career now? How can you adapt to a world in which AI powers, everything and continually alters consumer expectations of your brand?

[00:01:10] Let's hear what Mathew has to say. But before we dive into the conversation, let's take a moment and thank one of our sponsors. rasa.io. rasa.io is the only AI driven, smart newsletter platform designs organizations of any size can get more from their email programs by delivering relevant, personalized content to each individual on their email list.

[00:01:32] By sending through the rasa.io platform, everyone gets a custom version of the newsletter. The rasa.io platform can combine an organization's original content with third-party resources that are relevant and trustworthy. The result is that you have a compelling vehicle to remain engaged with readers on a high frequency basis.

[00:01:52] rasa.io also captures a great deal of insight data that you can utilize for marketing sales, product development, and more. Visit rasa.io to get started today. All right. Welcome to the Marketing AI Show. I'm joined today by Mathew Sweezey Director of Market Strategy at Salesforce author of The Context Marketing Revolution and host of the Electronic Propaganda Society podcast, which is insane by the way.

[00:02:20] I loved what you did with that. So welcome Mathew. Great to have you here.

[00:02:25] Mathew Sweezey: Thanks Paul. Yeah.

[00:02:26] Paul Roetzer: So Mathew and I were just like, we got chatting before this started and we were like 10 minutes and we're like, Oh my God, we should probably just start recording. Cause we could talk about this stuff of all day.

[00:02:36] And what triggered this one, Mathew was actually going to be a keynote for our, our Marketing AI Conference in 2020, which we obviously had to cancel. And he was going to be talking about the post AI consumer in that, but then he shared something on LinkedIn recently that was like, AI is going to be essential in your role.

[00:02:54] And here's the four reasons why. And so I like kind of went over and checked out. This AI is calling posts that he had, and I was like, this is just a perfect topic to talk about. So that's what this podcast will be about. But Mathew, I want to take us back a little bit because. One of the things you and I touched on before we hit record on this is this idea that consumers use AI dozens of times every day in their life, but they don't, they don't seek that technology and that you and I share this belief that moving forward, AI is just going to be seamlessly, integrate everything you do, but you were the number 13 employee at Pardot back in 2009, a marketing automation tool that got acquired by Exact Target for 95 million, and then Exact Target got bought by Salesforce.

[00:03:36] You've been in the marketing automation world since the early days you wrote the book on marketing automation for dummies in 2013, take me back to 2009. Are we even talking about AI back in 2009? Are we thinking about it as an industry?

[00:03:51] not at all.

[00:03:52] What was marketing automation in 2009?

[00:03:54] Mathew Sweezey: Revolutionary 2009. So let's go back to 2000 and at first off set the landscape, right?

[00:03:59] If you look at where we are in the marketing landscape, just go to Scott Brinker's map. MarTech landscape in 2009, there were 150 marketing tools. We were one of three, or I think at the time they might've been five marketing automation vendors, only three of those Marketo HubSpot part HubSpot wasn't marketing automation until 2015, 14, 15.

[00:04:20] They used to say marketing automation ...they didn't believe in it. So. It was a revolutionary concept. And when we were talking with marketers about this and let's go to the core, basic function of marketing automation, Boolean logic, if then do this, if email's not opened, wait three days, send this next email.

[00:04:40] If a person looks at this webpage, score them with this score. So we know who to talk to about what super basic functionality that basic concept was revolutionary in 2000. Go to modern time where we are now, first off, the landscape is 50 times larger, right? There's now 8,000 plus tools inside of the marketing [00:05:00] technology landscape marketing has become the largest IT budget of every organization, eclipsing the actual IT budget for the IT department.

[00:05:08] And so you have this radical different world, not to mention the world has changed as this has happened. Right? So like, you know, we'll get into this in a little bit, but we have to really understand that. Marketing was so basic in 2009, we were just doing email blast. People were still buying lists. I had to tell people, please don't buy lists.

[00:05:25] And we literally had to have policies put in place that you couldn't buy lists. And you had to check a box that you were importing opted in context, that you did not buy these. Like, we didn't have to put that into like 2013. But that's still what people were doing in 2013. So the concept of like where we were with super basic digital was super nascent.

[00:05:44] Social was, you know you know, post super basic, everything was super basic compared to where we are now. So let's just go with 2009. We were a baby.

[00:05:53] Paul Roetzer:  Yep. So Exact Target gets acquired by Salesforce for two and a half billion in June, 2013. [00:06:00] You're still there for that acquisition. Right. You're part of, and that's around the time you wrote Marketing Automation for Dummies.

[00:06:06] Mathew Sweezey: Exactly, exactly the time timer, Marketing Automation for Dummies.

[00:06:09] Paul Roetzer : So I wrote Marketing Performance Blueprint around that time, 2014. And I remember going through these exact scenarios in the book you got Marketo you got HubSpot IPO's in 2014, you got the Exact Target acquisition for two and a half billion.

[00:06:22] So we're literally talking about billions of dollars in valuation, billions of dollars in acquisitions and mergers and not a drop of artificial intelligence within the products. That are being valuable. So one of the things I always say within the, for our audiences, you haven't missed this yet. AI is moving really, really fast, but we are talking about seven years ago as an industry.

[00:06:46] The biggest companies in the industry still weren't even thinking about it. They didn't have machine learning, engineers dedicated to marketing and sales. They, they were thinking about human powered rules-based automation.

[00:06:57] Mathew Sweezey: And you know why, you know, while we were, let me, this is an important fact. One of my friends had a startup and this was an AI based startup back then, what it was going to do was automatically create websites, dynamically, create websites on the fly for infinite personalization.

[00:07:10] You know, they failed the reason they failed was because marketers wanted complete control over everything the customer saw. They did not trust a software to create the best experience. That'd be.

[00:07:22] That's changed 100%, at least for the smart marketers.

[00:07:26] Paul Roetzer: Okay. So you're an evangelist for Salesforce after that acquisition, 2012 to 2015.

[00:07:31] Are you talking about AI as an evangelist? I, you and I met on the speaking tour. I mean, that's kind of how we got to know each other.

[00:07:37] Mathew Sweezey: Yeah. Well, that's definitely how we met. I don't think I was talking about AI in 15. I think it probably took two. I've probably only been talking about AI to serious level for probably three years.

[00:07:48] Paul Roetzer: Okay. Yeah, because then you move into the principal of marketing insights from 15 to 20. And then you just in January of 20, moved into this director of market share. Now keep in mind for everyone listening. Like Salesforce has been one [00:08:00] of the leaders. I mean, they've spent, it has to be North of $10 billion buying up

[00:08:04] AI tools like they've bought all these things kind of mashed them into Einstein. So again, Mathew is at what would be considered by many to be a leader in the AI space, within the marketing sales industry. And even there, it's only been in the last three to five years where AI has really surfaced to the top of minds and where you started to see within your research and even in your writing and speaking to where it's now kind of landed to where we're in this, like.

[00:08:32] Post AI, consumer, where you're already talking about post AI. So why don't we kind of transition into where are we today? Like what, when you look at the landscape now, how do you think about AI?

[00:08:43] Mathew Sweezey: Well, there's different landscapes. Let's set the stage first off. There's a business landscape. So when we ever have a conversation with AI, we always frame it from me, the business employee, how does AI affect my business?

[00:08:55] That's usually how the conversation is always lit. It's never looked at from the consumer [00:09:00] standpoint, which is why I wrote about the postdoc consumer notch. I coined the term posted in my book with the postdoc consumer, and we must realize that here's what happened. As soon as 2009, theoretically, I wrote about this, proved it out mathematically 2009.

[00:09:14] Boomers became the largest creators of content in the world, surpassing businesses. And number two, the largest creator of content or their personal devices. Think about the notifications that you receive. Think about a Fitbit and how it motivates you to take five more steps. That's all personal content.

[00:09:29] When we look at that, now we start to realize that organic reach on social is less than 1%. The reason that is is because there's so much content that you are now competing, right? Mark Schaefer was about the content glut, right? We have this massive amount of content. Well, that is a problem for a human.

[00:09:44] So how do the platform create the best experience? Well, it leverages AI to create a contextual experience, not a chronological experience. And so now AI is sorting anything that a consumer touches that is digital it's completely controlled by artificial intelligence period. 

[00:10:00] So what a post AI consumer is, is a consumer never asked for AI, but every consumer that touches anything digital before they actually get to the end result is being filtered through artificial intelligence.

[00:10:12] Because that is the modern world to create the best possible experience. Right? So that is the post, the non-consumer they never asked for it, but everything they touch in digital world is artificially intelligent, artificially controlled by AI. Then you look at the other side. So that side is very advanced, right.

[00:10:27] You know, Netflix, Amazon on Facebook, Spotify pick, pick your points. From the business standpoint, that's a different landscape. That is a, how does my business operate function produce create. Now, if you were one of those companies, we listed, you are a very advanced AI practice because you know how AI produces the best outcome for business, because that is what you were selling.

[00:10:47] You are selling an experience. You have understood how to optimize that experience and leverage AI to do so. You've also learned to leverage AI to create better revenue potential, right? Amazon 13% of Amazon's revenue is [00:11:00] from their ad server. What do you think runs the ad server right behind the scenes.

[00:11:04] That is a lot of fancy technology that is saying what ad goes in, what place, how do we optimize price? How do we do all these different things? Right. We're starting to see AI in all these different formats. Netflix, you're going to see different images. For the different shows based on what AI thinks is going to be contextual to you, to get you to watch that film or show or TV show, whatever.

[00:11:24] So the landscape is very different. Every consumer touches it, not every business has it. Business is having in different applications. We were talking about this earlier. So we survey and ask everybody, which we do. I think the number is 84% of marketers use artificial intelligence, but we must be very clear when we use that number, because that is a, it's not a.

[00:11:43] AI is like a sliding scale, right? Every marketer uses AI because every tool that you have has AI baked into it. If it doesn't, it's going to right now, the marketplace is as follows every there's there's tools, there's tools that can be up-leveled within [00:12:00] additional purchase of adding artificial intelligence to it.

[00:12:03] And then there's tools that are brand new that come out with AI empowered in them, which you're not even asking to buy it. It's just. They know that the best way to get you the best outcome is to give you AI automatically. Hence, you're going to stay with my tool because it creates the best outcomes for you.

[00:12:17] You didn't ask for AI, you asked for better outcomes and that's what they delivered. Right? So that's where we're going. But right now, most people are in this process of. I've got a standardized technology set. I've invested in all these tools. I've got to maximize those values before I asked for any additional money where I can then start to add an AI that's even if they really realize they need it.

[00:12:36] So I'm going to stop with that because I would just, I can go on and on in different landscapes.

[00:12:41] Paul Roetzer: Yeah. And it opens up so many paths to go down. The one that I think is interesting. So one of the ways we talked about this at the end, the marketing Institute is it's just smarter technology. Like at the end of the day, you buy marketing technology to solve a business problem or achieve a goal.

[00:12:57] It's the only reason you would buy it or only reasons. [00:13:00] So if you're thinking, Oh, okay, I'm missing the boat. I got to go get some AI. I keep hearing it is really important. No, like. If you do email, you just need an email solution that helps you be a better marketer, helps you do your job more efficiently, helps you create convenience and personalization for the consumer without invading their privacy.

[00:13:21] And as Mathew is saying, The only way to do that is with AI. So you can go to your existing solutions and say, are there smarter ways to do what we're doing? Is there an upgrade option where I can get some machine learning that helps me predict things like I keep hearing it's supposed to be helpful, or you can go find tools that are being custom built to do very specific tasks within marketing, but that kind of spectrum you talked about of, you know, there are these companies that are building specific tools even then, though, just because they have a doesn't mean it actually works or they've proven their AI is any good. And I feel like that's a challenge. Yeah. That's a challenge we're having is as a consumer, like how do I buy [00:14:00] AI tech as a marketer?

[00:14:01] Mathew Sweezey: Well, there's a trick because a lot of companies are going to say they have AI, but really just ML.

[00:14:06] Right? We don't even have the market as consumers of this technology, even understand. If you ask a marketer to define the difference between ML and AI, I guarantee you that most of them can't define the difference. And that's a problem because a lot of the companies that are marketing AI really it's just ML.

[00:14:20] Right? So it's not true AI, right? So we have to be also as an industry, really be clear on what is it actually redoing and understand the difference between ML and AI. So when you actually are buying company, know what you're going to be getting. Yeah, the way we talk about it and I'd be interested to hear your take.

[00:14:36] Paul Roetzer: And if Salesforce even has a point of view and whether it's the same or slightly different than how you look at it, but what we look at is generally speaking, AI is sort of the umbrella term for these tools and technologies that make the machine smart, like Demis Hassabis from DeepMind has my favorite definition of AI, which is the science of making machines smart.

[00:14:52] They don't not to do anything on their own. The primary subset of AI is machine learning and that is making predictions about future [00:15:00] outcomes based on historical data. And then it keeps learning and evolving. So what we see though is ML is the abstract, the more abstract thing. People think it's all data science, which it is, it's just math, but they think only data scientists and engineers can understand machine learning.

[00:15:17] So they, they kind of shy away. But if you think about the branding of the industry, it's Google AI, it's Microsoft AI, like they're running ads. Branding AI, they're not talking about machine learning because they're just trying to get people to understand that smarter technology exists. And you're going to have to give up your data to get it.

[00:15:34] Like, that's my perception of where we're at as an industry is the big guys are all just trying to make us as consumers comfortable with this AI, whatever it is. Because when I go into fitness, plus from Apple, it's going to say, Hey, we're going to watch everything. You click on every instructor, every course in this thing, because we're eventually going to personalize your experiences.

[00:15:54] Is it okay to have your data? They're only asking that because they're their machine learning is going to learn what I'm doing to [00:16:00] predict what I would likely do in the future. But like, that's the, it is a challenge. And I talked to lots of these vendors and they're struggling to differentiate themselves with the B2B buyer.

[00:16:12] Because they do have better technology, but the buyer doesn't understand machine learning. So like they can't talk about the machine learning to those people. And so they go back to what you said. They talk about outcomes.

[00:16:23] Mathew Sweezey: Well, yeah, from a standard marketing practice, you, you sell the sizzle, not the steak, right?

[00:16:27] You shouldn't sell features. You should sell benefits and outcomes. And the outcomes are, we can produce better outcomes. It doesn't matter what you're using behind the scenes. Right. Because then you're going to start to slice them. Well, well, here's my algorithm and here's why that algorithm, nobody needs to know that, right?

[00:16:43] Like, no, one's going to care. You don't want to know what goes into spaghetti solace. You just want to know it's good spaghetti. Like that's all that matters. And so we had to get to that place of outcomes and AI does have limited, like ML does have limitations, right? So it's like, let's go back to the use case of in the B2B space for years.

[00:16:59] There's [00:17:00] been an, and I'm curious to hear your take on this, but for the past few years, and I don't know if we do it as much now, but there was definitely a lot of technology that said, listen, tell me all of your customers. And I'll find, I'll tell you who the new customers are that you should go focus on.

[00:17:12] Right? So using AI and NML to go find it. The problem with that is. Is, it was very hard for the technology to say, okay, how to go into new markets. It couldn't identify new markets. You could only identify existing customers inside of the marketplace. You already selling it because he couldn't make the connections outside.

[00:17:30] You know? So we have to understand there's limitations. That will be solved by it. True AI. Right. Whereas ML can't go outside of that.

[00:17:38] Paul Roetzer:  True AI means that like, Intelligent like true general intelligence where it finds problems to solve because where we're at today. And however you term it at the end of the day, what it's enabling us to do is make predictions about outcomes and as you're saying, it only can make predictions based on existing inputs, the data it's taken in in the past. So it's not [00:18:00] going to, it solves the problems you tell it to solve. And then as a human, you have to figure out if that's a good solution or not, but you have to give it the problems to solve. So like, it can be semi creative in its approach to things, but you have to give it the guard rails of what, what to solve for.

[00:18:19] So, yes, it's not going to find the anomalies like. And this became a huge thing with COVID. So if you were making all these predictions about lead scores or about consumer behavior, if you're an email marketer, like, do I use emojis? Don't I do I use free in the messaging, like all of these things that machines could in theory, make predictions on by learning from thousands of emails you've sent or hundreds of thousands.

[00:18:42] Those almost became a relevant in April of 2020, because the consumer changed. The consumer wasn't going out anymore. They, they, you know, the things that, and the machine learning models couldn't have foreseen out ahead, they had to now learn from an entirely new data set of how this consumer behavior [00:19:00] was evolving.

[00:19:01] Mathew Sweezey: And that, that is a deficiency in machine learning, but it doesn't mean it's not valuable. Yeah. And it's also a deficiency in how marketers think about data. Marketers are going to have to rethink their data strategy, right? Their data discipline freshness has to be a key critical component of your data strategy for this exact reason, because you have to realize that if you, and so let's pivot this conversation really quick.

[00:19:23] Talk about that. The Post cookie future. We have to understand that first party data is the key to everything that we're going to be doing in the future. Right. That is the number one thing.

[00:19:31] Paul Roetzer: Can you explain what you mean by that real quick, but just so people understand like how the world exists right now on cookies that the automation is enabled by cookies, but when we can't use them, And why we aren't able to use them for a second?

[00:19:43] Mathew Sweezey: Yeah. First party is data that you collect from somebody such as someone comes to your website. It's your own property. You're allowed to collect that that's first party. There's the term zero party, which is a subset of first, which is essentially [00:20:00] anything that someone explicitly gives you such as I click this box, I filled in my name.

[00:20:04] I give you explicit permission and I filled out a lift, right? Given data. And then there's second, which is shared first party data. So I'm Amex your Delta. We go into a partnership. We share data that second party, third party is me tracking somebody else on some other person's site that is what's going away.

[00:20:22] So third-party cookies are going away due to privacy concerns, right? This is the marketplace regulating itself before the government does rec Coney and measures. Exactly what's happened. You then move forward and say, okay, how does that then change everything? It changes everything. Ad tech marketplace goes away, right?

[00:20:39] You're going to have everything now inside of walled gardens, the concept of multi-asset multitouch attribution goes away because you can't do it. If you can't look at where people are going in places you don't own. Right? So th th this is going to radically change so many different things. What then do you do in that new world?

[00:20:55] Well, you have to rely on first party data. And so what you're going to do is we had this big shift [00:21:00] to content marketing. Well, that's going to get accelerated in content. Marketing is going to get reformed and the first party data collection, that's exactly what we expect to be happening, which means that you're not just going to have content you produce.

[00:21:10] You're going to be hosting free wifi. You're going to be creating data at a dating apps. You're going to be doing a thousand different things that are out there. Designed to collect very specific data from very specific people that it's continuous, which brings in fresh out of this. It goes to the freshness team because you need to have a continuous.

[00:21:27] Stream of new data to then run these programs run on the day. So that's what I mean by the freshness and that data discipline and why first party data is so critical because it has to be fresh for this data, the AI to know in real time, what am I looking at? And then this then brings up lots of other questions, such as, you know, how much data do you need, how good is data?

[00:21:46] You know, do we decay this out over time? Like we have a score that an indicator. There's going to be a lot of Data Scientist can be a big one thing anyways. That's just kind of, no, that's awesome.

[00:21:55] Paul Roetzer: And a really practical example for people is like I was doing a webinar recently with R.J. Talyor at Pattern89 and they have, they've built an AI tool to predict the success of ad creative on like Facebook and Instagram before you run it.

[00:22:09] So they learn from all the performance data from, you know, millions of ads and then you upload creative variations. And it'll predict which one is going to work before you spend a dollar to run it. But their data shows that the average creative runs its course within 10.4 days. So like, if you're an advertiser today, you're going to go back like, Hey, what worked four months ago?

[00:22:31] When we ran the last Facebook campaign, did we have people in the image or was it just like characters or was it computers or what was it? Was it just text? Let's just do that again. Well, that ad alone was old. And over by the time you finished three months ago, more or less recycling it to be the ad you run now.

[00:22:50] Mathew Sweezey:  And so just this idea of recency and this idea of like prediction it's. I don't know what you just did. So there's a whole new concept that me and Brian Solis are working on. It's called [00:23:00] back to advertise and get ready to see a lot more of this moving forward, the data that you just shared explicitly states, why we have to think about fast advertising.

[00:23:08] If creative is dead in 10.4 days, the average consumer sees more cons than they've ever seen in their entire life. The question is how do we stay relevant at scale? For brands that specifically rely on mass advertising methodologies to have a relationship with their consumers. I hate to say it, but as much as we have people in massive brands, I'm just going to pick on toilet paper, toilet paper brands want to have a personal relationship with their consumers for obvious reasons, their brand.

[00:23:32] They want to have a personal relationship. I don't want to have a personal relationship with my toilet paper brand. They need to do one thing. They do one thing. Well, that's all I care about, right. And so the question we have to think about Dennis moving forward is then how do they then stay relevant?

[00:23:46] How do they build these relationships? Well, advertising is the method that they're going to use. It's push button, it scales, it gets them that relationship. The problem is to exactly what you just said is the traditional method of advertising runs on what I would call a [00:24:00] traditional fashion model, which is we have four seasons.

[00:24:02] We're going to have four major campaigns. We're going to run a year, but look, what happened to fashion. It went to every week, it's a new fashion trends. Fast fashion advertising is going to have to keep up and exactly what Pattern89 is doing is what we have to rely on. AI is going to be that empowering method, which says, all right, marker.

[00:24:19] Here's what's happening here. Here's your sub audiences. Here's what's actually what they care about in real time, by the way, that's going to be curated by IDI. Then the next step is going to say, okay, here are the things that we think you should talk about. Here's the messaging that we think you should come up with.

[00:24:33] That's going to be powered by AI you're then going to refine that, add to it. Then there's going to go into two different types of tests and you're gonna have a pretest thing of let's test that message in real time, that's going to be inside of a lab format. It's not going to be public. There's going to be posts, which is going to say, okay, now we've got the messaging.

[00:24:48] What is the actual advertisement that's going to get put out into a lab? Were they going to be able to test how people react to that? Whether it be with emotions, whether it be facial recognition, whether it be by clicks, whatever that's going to be driven by [00:25:00] AI, then there's gonna be the programmatic execution of that advertisement.

[00:25:03] That's going to be done by AI. Then there's going to come the feedback loop, which is, Hey, listen, we saw all the ads you produce last month. And we realized that any time that you have someone smile, you have a click through rate increased by 10%. That's going to be done by AI. Right? So when we start talking about all these layers of AI, I made the prediction two years ago that by 2025, all marketers are going to become AI reliant.

[00:25:27] Right. And that means that you will have a very hard time competing without artificial intelligence because artificial intelligence is going to drive efficiencies so efficient. I love it. So what does the human still do? And that scenario you just listed like 10 elements of just advertising, where they guy's going to do it.

[00:25:44] Paul Roetzer: What's the human still do in 2025?

[00:25:46] Mathew Sweezey: Oh, all of the stuff that humans do well, AI, can't go into a meeting and say, this is why we need to have this, that AI can't go create the strategy of what we're going to do, where we're going to go, how we're going to do it. AI can't, there's so many things that AI can't do right?

[00:25:59] The things that we've relied on humans to do for so long, AI is going to replace and be more efficient. Now the creative element, AI is going to be able to tell you what works. I don't really believe that AI is going to replace the creatives. The creative, going back to the initial problem that we just said earlier of, if we say AI, tell me new businesses, I should go after it can only do that inside of the data that you've given it.

[00:26:22] Initially it can't expand and create new theories and new eyes. I know you got a conversation yesterday about a total entirely new vertical that could open up enormous possibility. It's not going to know that information, unless you say, Hey, go analyze this vertical for me. Tell me like what you learn about that.

[00:26:38] It's not going to know to look there though, unless you tell it to look there. Yeah. So it's going to be a tool that's going to be leveraged and it's going to be an enterprise scale. Everyone's going to be leveraging these things cause they do all kinds of things. But specifically when the marketing expect by 2025, you're going to be reliant on AI, at least in one of those five categories, at least in one I loved in your AI is calling.

[00:26:58] Paul Roetzer: We’ll attach the deck and we'll include the [00:27:00] link in the show notes, but you talked about how Mars used facial coding version of AI to determine their best ads for M&Ms. Can you explain that for us? Because I think it brings a very tangible example to that list you just gave all the things they can do of a brand that's actually using some really advanced stuff.

[00:27:17] Mathew Sweezey: Yes. So this is a case study. So Mars essentially said we want to be able to no, they're using advertisements to create an elicit, a specific reaction inside the human. They want to create a very specific relationship and they're using AI to see in their ads. What emotion does that elicit and the person watching that advertisement.

[00:27:36] And there's very specific things they want to listen. So they use AI and facial recognition. So they play their advertisements in front of somebody while using a camera that tracks their facial expressions, but then notices how their face changes, which has been the representation of emotion. They're able to then say, we know this ad is the most effective at delivering the emotion that we want in our target demographic.

[00:27:57] Then that's the ad that they then show they have [00:28:00] their own proprietary ranking system of like what they're looking for, but that's where they're using AI. Now. Caveat. That technology is effective, but it's also questionable. There are questions. You'll see both sides. You'll see a vendor saying that this works, you'll see Mars using it.

[00:28:15] You'll also see vendors who are competing against that, who will say that may not be the best. There may be other methods. So all we're saying here is there's lots of different ways that this is going to be used. We don't know what the best method going forward is, but we do know. The AI is going to be able to identify things that we can identify as humans.

[00:28:33] And it's going to make us make, help us make better decisions regardless, moving forward. Yeah. You and I could just talk for hours. Um, I'm going to go back to your LinkedIn post, which is again, what led to this. I'm gonna, I wanna just talk about these four items and I want to talk about your book for a minute, and then we'll, we'll kind of wrap up.

[00:28:48] Paul Roetzer: So your LinkedIn post and I'll just kind of read the first part, says AI is a major part of the future of marketing full stop. This is obvious, but there are some ways that I will change things you may not have considered. And then you listed four. So I'm just going to give [00:29:00] you each of these four and you just kind of give me a little narrative.

[00:29:03] Deep fakes may be a positive, what is a deep fake, and I've only heard about them negatively. So why, what is your point of view there? Yeah. So deep fakes, as we know them is any time that AI sakes or a person has AI take a person and an image and have them do something that they didn't do. So it didn't manipulate the image.

[00:29:22] Mathew Sweezey: It could be Barack Obama riding a horse, and he's never been on a horse. Right. So like, it could be anything that you want,

[00:29:27] Paul Roetzer: or it could be your CEO saying something that he or she never said that could create a crisis plan. Yeah.

[00:29:34] Mathew Sweezey: Yeah, it's essentially, if anything, the problem that we all see is the bad parts of this, but there are good parts for us as a marketer.

[00:29:41] And those are some very basic things. International translation is the easiest example to consider. Right? So David back on, there was a great example, David Beckham, where he shoots an ad, talking about why you need to wear a mask. It's instantly translated into 17 different languages. David Beckham space.

[00:29:56] Then the mouth changes to match the language in each one of [00:30:00] those. So it looks like he's speaking each of those languages. So infant language translation, imperfect language translation. So anyone that runs global brands knows the issues of global translation. It's very difficult. AI can solve that very easily.

[00:30:12] There's another concept in the same thing, which is what if we were able to deep, fake the person, watching the ad and put them in the ad. That's a crazy thought. Right? So rather like that is ultimate personalization of like, rather than like, you know, you watching two people sit intelligence for the Viagra commercial, you and your wife sitting on the hill and the Viagra, he gets credit Viagra, if you use that idea. So those are some easy examples of deep fakes.

[00:30:40] Paul Roetzer: Shopping on the edge is the next major wave of commerce. What is shopping on the edge?

[00:30:45] Mathew Sweezey: The edge is defined as any place that is not a traditional brand or retail site, Instagram shopping, Pinterest shopping. These are great examples of edge.

[00:30:53] Essentially what you're looking at is the edge is going to open up all kinds of crazy things. Full funnel experiences is really what AI is going [00:31:00] to open up on the edge ad to be targeted to anybody. Then that ad can be completely immersive by AR and VR. You can then go in and you can try on the products.

[00:31:09] You can demonstrate the products, um, using the Ikea place app, or you can actually put the furniture in your room. See what it looks like. You can then go through a chat bot, which is an AI driven chat bot and have a conversation in real to can check inventory, purchase everything that happens all within the same moment on the edge, completely compressing the buyer's journey tweets single moment, that was targeted to an individual that is on the edge, no human interaction.

[00:31:32] And by the way, those things are highly effective and efficient. Did the AR with a TV, my son got overzealous and threw something into our. Nice 55 inch screen TV the other night. And so I went on the best buy site and I'm shopping and I was like, Oh cool. I can visualize the TV in my room. And I did that.

[00:31:50] Paul Roetzer: Pop it up there. Oh yeah, it looks good. Okay. Fine by it. All right. Number three. How do you, how you do your job?

[00:31:56] Mathew Sweezey: So we discovered this one in the past, like 10 minutes of conversation. AI can change [00:32:00] every element of your job moving forward. You know what you're going to do. Here's a really easy example.

[00:32:05] You're going to have your own personal box. That's going to direct your daily work. You're going to walk into the office and your boss is going to say, Hey, listen, last night I was looking at the website and I've found that there's a new segment that you haven't identified. I looked at our current ad spin.

[00:32:17] I believe if we shift this percentage of ad spend to this new demographic, we'll be able to see this return. Do you want me to do that click yet? That's what's going to happen and how it's going to dictate our tasks and then moving forward. All right. Last one, AI opens up full funnel experiences. So I just, I, I hit that one cause I forgot what I wrote before I wrote about, but let's play with one other one, which is, let's talk about AI as native product placement and a new idea.

[00:32:43] Paul Roetzer:  I love that. Well, I've seen you talk about that one. I love it. All right.

[00:32:45]Mathew Sweezey:  So here's the concept. Native advertising is essentially the content that we are putting something in line that looks native to the format, right? So that inside of Facebook, that means it's a post inside of an art inside of a media property.

[00:32:57] It looks like it's the native article. So there's lots of [00:33:00] concepts. But now let's go to, let's go to television and mass advertising, right? When we have OTT, which means over the top television where we know there's a logged in user, we can identify who they are. AI now allows for native product placement inside of television shows in real time.

[00:33:15] So rather than everyone watching American idol and seeing a Dunkin Donuts cop or a Coca-Cola. I can say this demographic sees seven up. This demographic sees Fresca, this demographic chief Cambodia. Right? So it just depends on what you want to do. And that's going to open up this radical new world where everyone will see a different TV show.

[00:33:35] Were all the products inside of that show, even in the background are going to be programmatically ad based and AI inserted. So it's almost like when you're watching like a MLB playoff game and they've got the green screen behind the batters, and I'd never even thought about the fact that different markets might be seeing different ads, but you're saying that could be like a cup could basically be built into the production to function like a green screen in essence.

[00:33:58] Paul Roetzer: And you could surface whatever ad you want [00:34:00] or whatever product you want in that. Wow. All right, man. So what, uh, what do marketers do to get started? I mean, we just talked about some crazy, crazy stuff for five years from now. Like if, if you're a marketer and this is, you're just listening for the first time and you're just starting to kind of get into AI and you are going to be intentional about trying to find some solutions.

[00:34:18] What's your recommendation for where they start?

[00:34:20] Mathew Sweezey: Step one, take a deep breath. Um, we've just covered a lot of stuff that could freak you out and it could really say, Holy crap, I am so far behind. I'm never going to get there. The stuff we just talked about is radical. It's super advanced. Like these are not going to become mainstream for years to come.

[00:34:36] Right. But we know they're coming. We already see examples of them. Right. So don't get freaked out too, is having a very clear, specific goal that you're trying to accomplish and then work to accomplish that goal. Right. So it doesn't, you don't need to go like you don't need to go home while focused on a specific task that you're trying to accomplish and then find the best way to accomplish that task, you know?

[00:34:56] And then when you're going through the evaluation, Maybe educate [00:35:00] yourself on what is actually behind the scenes. Is this ML, what kind of the mail, what data sets am I going to have to have, and then walk into these things, but start testing now, right? Don't have to do it across your whole stack, but you do need to start investing and understanding where it plays and how it plays.

[00:35:14] And most tools you can just simply call your vendor and just flip AI on, um, you know, all of our products are enabled by Einstein, Einstein, AI. So it's, it's either depending on the product's gonna be built in or it's going to be an adult, just depends. You can pretty much flip the eye on anything right now.

[00:35:29] Paul Roetzer: All right. So before we get into the, the last element of the rapid fire, tell us about The Context Marketing Revolution, which came out earlier in 2020. Tell us a little bit about that book. The best marketing advice that I learned from this book was don't publish a book in a global pandemic number one key thing.

[00:35:48] Mathew Sweezey: Right? So what The Context Marketing Revolution really means is it's a, it's a theoretical book with practical advice, but theory Is the very definition that we know of marketing no longer exist [00:36:00] and it no longer exists because the media environment that we operate in no longer exists. So I approved that we entered what is called the infinite media errands specifically. It happened in 2009. We already talked about what that means anecdotally, because now consumers that the largest creators of content second largest cruise in concert are their devices. Well, we must realize is marketing is a game and the rules of the environment dictate the games that we play.

[00:36:22] So old games that we thought about such as sex sells, right? Message right person, right time. The apex of direct marketing is one-to-one right. Those were games that we played based on the environment that we operated in. In a new environment, we must think about new games that we play. So really what the whole context marketing revolution is about is saying, we must redefine what marketing means inside of our organization.

[00:36:43] There's a big conversation about digital transformation that's happening. If you don't transform the very definition and idea of marketing, I'm not saying get new marketing tactics and techniques, I'm saying scrap. The idea that you have of what marketing is. Re-imagine it to fit the new world and the new [00:37:00] reality, and that is then what will succeed in the future moving forward.

[00:37:04] Paul Roetzer: So definitely, definitely check that out. Uh, all right. So we always end with a rapid fire. Before we go into that, I want to tell you a little bit about one of our sponsors MarketMuse. MarketMuse Suite and AI powered content intelligence and strategy platform analyzes millions of articles on demand, uncovering gaps and opportunities for better content.

[00:37:25] The output empowers marketers to craft high quality content that their audience loves and search engines. Reward MarketMuse uses AI to accelerate content planning, creation, and optimization market. MarketMuse has a team of content and AI veterans using machine learning that enables clients to gain authority in their topical domain.

[00:37:44] Receive 20% off select packages for your first year with code Academy20 that's Academy two zero. Visit MarketMuse.com to get started. All right, Mathew, are you ready for the rapid fire? I have one [00:38:00] in here that we haven't asked anyone else, because I know you were at an adventurer, so we're gonna, we're going to end with a fun one personalized to you.

[00:38:08] All right. First, first question, voice assistant. You use most Alexa, Google assistant Siri Cortana. Don't use them.

[00:38:15] Mathew Sweezey:  I'm an Apple guy.

[00:38:16] Paul Roetzer: I'm a Siri guy, but if I actually need an answer to a question, I have to go to Google. Like just very often, won't get it. So I'll have, I have Google assistant app on my phone too.

[00:38:25] And if it's like, we're on the way to school one day, my daughter is like, how far is it to the moon? And you're asked Siri and the answer, it was like, it was like, not that useful. Yeah. As Google. And it was like, envisioned this. And it's like, you stack it up. It's like, Oh my dog. My dog is like, Oh, that was a much better answer.

[00:38:39] Ask Google next time. Okay. More valuable in 10 years, liberal arts degree or computer science degree.

[00:38:46] Mathew Sweezey: Ooh, that's a tough one. I think that really just depends on what you're doing, but in terms of like, if you, I think they're both inherently invaluable. I think they do. They're two sides of the same 0.1 is how do I think about these things?

[00:38:57] The other is how do I practically do these things? So I think they're [00:39:00] both inherently valuable, but I think those are probably more valuable than just general business degrees moving forward. All right, net effect over the next decade. More jobs eliminated by AI, more jobs created by AI, or it's not going to have a meaningful impact on jobs.

[00:39:14] That's always the million dollar question of automation. AI would not do it's more jobs. Well, I mean, we've seen this from studying everything from the fifties. When we thought that basic automation would, would take jobs away, it just creates more jobs in different aspects, in different arenas. The hard part is that, you know, there is a really good example that I was hearing the other day.

[00:39:34] We're worried about fat finger jobs. Right that finger jobs. If you've never had a family member who was a manual labor, you may not know, understand what fat finger means, but anytime you're working with your hands, constantly, those muscles get really big and you have these big burly hands. So that's the concept of tapping the job.

[00:39:49] When we start thinking about all these things, there's a whole infrastructure that has to be created. That is physical, right? There's gotta be wire. It's going to be fiber. There's gotta be boxes. There's gotta be all these other elements inside of this [00:40:00] world that are those jobs that it does take to support those things.

[00:40:03] Then there's the intellectual work of how all these things happen. I think a better, better question to ask is where do those jobs take place? Because now if we think about there's no physical barrier to these things, the question may be is where are those jobs located? Right. Not what jobs do we eliminate?

[00:40:19] Paul Roetzer: It's hard for people that AI is abstract to a lot of people. It's really hard to see around the corner and envision it. I understand that people are afraid of it. I understand. They worry, but there, there are enormous potential opportunities that it's just hard to see them coming. All right. Last two. What does an AI agent win first or at least share with a human, a Nobel peace prize, an Oscar, a Pulitzer, or none of them?

[00:40:45] Mathew Sweezey: Oscar.

[00:40:45] Paul Roetzer: Yeah, I'm a Nobel peace prize guy. I think scientific advancements.

[00:40:50]Mathew Sweezey: I mean, you can make an argument and I could agree.

[00:40:52] Paul Roetzer: I know, see too. Uh, it's almost like, which is the more progressive thinking body that awards them. That would [00:41:00] take the all right. Last one only time I'm ever going to ask this. This is only for you.

[00:41:04] Favorite adventure. So in your LinkedIn bio, it says outside of the office, I'm an avid surfer climbers, snowboarder mountain biker. Do you want to give me talking ass about my last adventure? I'm asking you about your favorite adventure either in the last year or all time.

[00:41:19] Mathew Sweezey:  Yeah. Favorite adventure of all time. So one of my best friends, his brother-in-law teaches in was Uganda. Now it's in Kenya. Um, and they had their parties on the weekend. So he's his own private four wheel. So we take a trip 10 days and we, uh, we for wheel across Uganda and privately and just have a ball, you know, Hey, camping with hippos, um, playing around with denial.

[00:41:40] Um, like it was. It was crazy. That was probably all time. Favorite adventure. When you wake up in the morning and your buddies, like. Did you know, there was a hippo literally outside our tent last night. And I was like, no, the sleeping pills not be completely out of. Yeah. I don't have the thinking bills by the way, they make you snore because you were snoring.

[00:41:58] I felt the hippo. [00:42:00] What year was that? Uh, that was just, I think, three years ago. Oh, wow. That's awesome. Well, next time we're together. I want to, I want to hear more. I'm not an adventure. Like I just, we traveled my wife and I traveled a lot, but I am not a huge risk taker when it comes to being in other countries.

[00:42:18] Paul Roetzer: So I live vicariously through people like you.

[00:42:21] Mathew Sweezey: No worries. The funniest part was before we went to bed, we went to bed with the sound of drums. No joke. There was, I don't know if it was like some tourist drum circle or if it was like some, some native populations on circle, but you heard them all night.

[00:42:35] And then we went to bed and then wake up with the story of the hip hop crazy world. That's awesome, man.

[00:42:40] Paul Roetzer: Well, any final thoughts for our audience? Just as they're trying to kind of. Wrap their heads around where we are in the next 12 months. And we talked a lot about the next five years, but just what they can do in the next 12 months to sort of get ahead of their peers.

[00:42:52]Mathew Sweezey: Just investigate. Just start asking questions and constantly learn. I mean, you're awesome. I listened to a podcast about AI for you're already [00:43:00] interested in just progress your practice, right? Maybe there's certifications. You can get, maybe you can start to have conversations of creating some type of a working group inside of your organization.

[00:43:08] That's focused on AI implementation. How can we solve new problems in new ways? There's lots of things that we can do, but I think that people that are listening to this it's keep continuing down that path that you're going down. Keep asking questions, keep investigating and keep pushing forward. I love it.

[00:43:22] Paul Roetzer: I just, as I tell people, take the next step, just read the next book, take the course, go to the conference. Like don't try and do it all at once. Just. Commit yourself to the fact that the industry is changing and you need to start taking those steps forward, best place for people to get in contact with you and find you online.

[00:43:40] Mathew Sweezey: I locked down LinkedIn, but Twitter is easy, but just reach out if you want to reach out and chat. If you've read my book and have a question, very happy to chat, this has been awesome.

[00:43:48]Paul Roetzer: We will do it again. I love it. I love talking about this stuff with him. All right. Thanks everybody.

[00:43:53] This has been the marketing AI show until next time we'll talk with you again. Thanks so much. Thanks [00:44:00] 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 Marketing AI Institute.com.

[00:44:10] 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.


Related Posts

[The Marketing AI Show: Episode 2] Microsoft Exec Covers Conversational AI

Sandie Young | January 20, 2021

You can now listen to the first four episodes of The Marketing AI Show podcast on your favorite podcast app. Read this blog post for more on what to expect in episode two.

[The Marketing AI Show: Episode 3] Can AI Create Content?

Sandie Young | January 21, 2021

You can now listen to the first four episodes of The Marketing AI Show podcast on your favorite podcast app. Read this blog post for more on what to expect in episode three: Can AI Create Content?

[The Marketing AI Show: Episode 4] The Dark Side of Social Media

Sandie Young | January 22, 2021

In this episode of the podcast, David Meerman Scott digs into the dark side of social media and considers the question: Is Facebook evil?