CMO Strategy: Marketing in the AI Era

Michael Krigsman: Today on Episode 817 of CXOTalk, we're discussing the chief marketing officer role and AI. Our guest is Greg Stuart, who is CEO of MMA Global.

Greg Stuart: We are a nonprofit industry body for CMOs, so we work with the world's largest marketers. My board chair is the CMO of AT&T. My previous board chair was the CMO of General Motors – just to give you a sense of it.

All trade bodies are really intended (all nonprofit trade bodies) to kind of a similar thing, which is, what does the future look like and what do we do to march the industry in that direction in an accelerated way? That's basically the focus of all of these.

Michael Krigsman: You run an extraordinary number of events around the world.

Greg Stuart: We have 800 corporate members globally. I have teams that operate in 15 countries worldwide. 

We run 35 conferences. Some of those are 4,000 or 5,000; there's one that's 6,000 people a year. Yeah, there's a lot to do this. 

I think that the work that we do, the initiatives, and some of the research and the insight development is the bigger deal. But yes, we are very big, as most trade groups, in bringing and convening people together, members together. Yes. 

Top challenges and focus areas of CMOs

Michael Krigsman: You are in touch with CMOs, as you said, from the largest organizations in the world. And so, you really have a good sense of what CMOs are thinking, what are they focused on, what are the challenges that they're facing. 

Greg Stuart: We operate in kind of a think tank mode. Our basic orientation is that there are huge gaps in marketers' knowledge about how certain parts of the business really work. 

Also – and this is the one that is even a little more discouraging – there are parts of the industry that many marketers believe to be true but just aren't so. And so, why the CMOs appreciate us is because we're truth-tellers. 

I have no axe to grind. I don't sell anything other than being a membership organization and trying to really make marketing and marketers better. 

It's, in essence, to raise the stature and gravitas of the CMO within the organization. It's to give them better credentialization. The way we go at that is, as you said (or we mentioned here), I work with the world's biggest CMOs. 

What's also interesting and a little bit rare, it's called "A big tent strategy." We also have all of the solutions companies, the tech companies (so, Meta at the highest levels), are sitting within the MMA. 

Google, Pinterest, Snap, everybody at the most senior level of those companies are focused on really trying to make the marketing industry a better business and better for everybody, too. Better for consumers, too. That comes a lot into the work that we do.

Michael Krigsman: Greg, as you talk with CMOs, what are the focus points and the concerns that CMOs are facing right now? I think that's a good way to establish context for our discussion.

Greg Stuart: Yeah, we operate in think tanks, so I've got a think tank that's focused on measurement and attribution. How do we really understand the return on investment of the decisions that we as marketers make? You'd be shocked at how bad we are about that, and I can give an example of some we do. 

We do a lot of work around marketing org and marketing org strategy. I'll give you a small example of that.

We have built a product. It's an actual product now. It started as a thesis that we worked on.

We can now tie marketing organizational decisions to actual financial performance of the company. We know how to boost sales versus detract from sales into the design of the org. 

Listen, Michael. I don't know. You do a lot of these calls with a lot of different C-suite executives. I've never heard that happen in any other part of the organization, and marketing has now advanced there.

We do a lot of work in data and customer experience. The underlying thesis of marketing has really been around brand. There's a big part of the business around brand.

We know from our research that the central focus of marketing going forward is around customer experience. That is today's winning strategy.

We're going to talk about AI here at some point as to the next iteration of those things, but those are really the three main areas where we play. 

Impact of generative AI on marketing

Michael Krigsman: What about AI? Let's shift into that. I guess the very first question is, from your perspective as you speak with these CMOs, how important is AI and generative AI at this moment in time to marketers? It seems like an obvious question, but maybe not so obvious.

Greg Stuart: We've done a series of the state of survey research with marketers lately. One of those was around the state of AI in particular. 

When we did that study, probably earlier this year, it would have been probably Q1. We actually believed about half of marketers at that point said they were scaling some type of AI effort of some kind in the beginning of the year).

We did a more recent study around generative AI and two-thirds said that they are just starting to explore. So, I think they're in it. 

Listen, Michael. I've done the research. This is kind of my job. This is the job of the MMA. 

I've done the research, and you'd be insane as a marketer not to be leaning into AI. It's the most important thing you should be focused on right now, in our opinion – based on our data and partly based on our opinion.

Adoption of generative AI among company marketers and CMOs

Michael Krigsman: Given that it's so important, where are we in terms of actual usage and adoption in the real world?

Greg Stuart: I still think that's a little slow. I get some marketers who are getting some pilots around some areas. 

Here's maybe a good way to kind of frame it, okay. We have a basic framework when we talk to marketers, and this is not to be oversimplified. It's not going to be that dramatic of a framework, but we basically say, "Listen. There are three things you would want to accomplish with AI," or anything new for that matter, but AI in particular. 

Where has it developed greater efficiency? Where can you improve the processes? How can you maybe somewhat even improve the product of marketing, especially the cost and expense of doing it?

Michael, I don't know if people understand. I was talking to a CMO of a big company the other day, and they were talking about a campaign they had launched that had 5,000 assets within it. 

Those of you around my age will maybe remember it used to just be a 30-second television ad or some variation of that. But the business of marketing has gotten incredibly complicated.

First, around efficiency. How do you bring those 5,000 assets to the marketplace, the content element, and push those out? 

The second part of it is, what are you doing to boost performance? Are there places? 

We've got a lot of research. We're going to get into that here. That's where I see some of the biggest opportunity. 

The other one is that I think that nobody is really... I don't know if people are working against it. I'm not sure we'd even know right now. But it's going to happen at some point. When does it become game-changing? 

What are we doing to reach for the stars so we can really move the needle? I don't know if we know what that looks like yet.

The first two are easier. But that's where I see most of the effort being organized around. 

Michael Krigsman: Managing content, being more efficient in creating content, for example, that's not exactly strategic if you're the CMO of a really, really large company. Now, it's important. But that's not going to change the game, as you said.

Greg Stuart: There's a certain amount of truth to that. But you also have got to make sure that stuff is done factually correct, which obviously AI, we all know, with hallucinations, has a little bit of a challenge. But also, too, that it represents the brand and it represents the brand voice. 

We were having a conversation. As you know, I have a podcast that I do now called Decoding AI for Marketers that a gentleman named Rex Briggs and I do. 

We had a guy on there the other day. It was very interesting. He said, "Yes, we have the AI trained," or generative AI then, "to develop content for us." 

He goes, "What we now need is a separate AI system that is validating that it got it right," which I thought was kind of funny. [Laughter] The machines checking the machines felt sort of as an unusual concept. 

I don't know that that technology exists yet. If it does, if somebody has brand voice. 

Actually, my friends over at Grammarly (if you're familiar with that tool), they've actually built into their next enterprise vision. I'm not an expert at this, so they can talk more about it. They built into their enterprise version the ability to make sure that communications going out from salespeople are reflecting the brand tone, brand voice, and some of the product elements.

I think we're on the cusp of that kind of thing, if I'm capturing what's going on out there. 

Michael Krigsman: There is early adoption that is, say, tactical.

Greg Stuart: Yeah.

Michael Krigsman: But maybe something strategic coming down the pike. Do you actually hear about specific use cases that are more strategic, or is that just still too much in the future?

Greg Stuart: I ran a panel yesterday for an event, and it was interesting. I asked people. I asked the panelist, which was either a group of CMOs or ex-CMOs. I said, "Where are you in thinking about AI?" 

You know what was funny about it, Michael? It was very strange to me. They talked about, "Well, we have to be a little bit careful. We have to manage the risk." Like, "We don't want to get too far ahead of ourselves." 

They give a very—dare I say—corporate cautious answer. So, I was kind of taken back that that's where they went first.

Then I said, "Well, how important do you think AI is? Give me one to ten," ten being "Oh, my God. Change the world like we've never seen before. One, "What's AI?"

You know what they pick? They picked 8.5, 9 maybe even sometimes.

I was kind of like, "Well, how can you--?" I kind of called them out on it, like, "Why are you focused on talking to me about risk. You sound like corporate people rather than people really trying to change the world or transform the company."

Now, what was interesting is that, as I listened to their conversation, I think they were getting pressure from their CEOs to go faster and they were trying to put the brakes on at some level in order for them to kind of catch up. It was my sense of what was going on in that conversation. 

Listen. As an entrepreneur, I run a company of about 100 people right now. I don't have that kind of caution. We are go; let's make it happen. That's what small companies tend to do.

This reflects the big corps. I think they still struggle a little bit with it, and I don't think we've got a good organizational system. But the MMA is doing that.

One of the projects we have going on right now is how you organize your marketing org department for growth through AI. And so, we'll be doing some research and capturing some frameworks around that over the next two, three months in trying to give that to marketers to give them an ability to move a little bit faster.

But now, I think, in answering your question, it still feels very early to me.

Obstacles to AI adoption in marketing strategy

Michael Krigsman: It sounds like there is a good amount of fear – I guess you could say fear of the unknown at this point – among CMOs.

Greg Stuart: This is the thing about marketers and kind of the thesis of the MMA is that nobody in marketing is prepared for AI. Most of us don't have technical backgrounds. Most of us don't even have good statistical or mathematical backgrounds to understand the complexity.

But you also see that show up in the rest of the business. I don't want to get too distracted from the AI topic, but there are basic parts of marketing that we are incredibly unclear about.

I'll give you a good example. I mentioned measurement attribution. About four years ago, I started to hear marketers talk, and I started to hear CMOs. 

They would say things. There was this tension in the organization around brand and performance, or brand versus performance.

If you talked to the brand people, they'd say, "Oh, those performance people. They're just focused on short-term goals. They have no appreciation for long-term brand, customer equity." 

If you talked to the performance people, they'd say, "Well, the brand people are full of rainbows and unicorns. They just make [explicit used] up. They don't really know what to do. They have no measurable goals."

There's a tension in the organization. What caught me about that situation, though, as we took a step back – this is typical MMA – we didn't ask the marketers what to do about it. I don't ask. I don't ask them how to make a faster horse. I never ask my members how to make a faster horse. I just don't believe in that.

It's our job to go back and say, "Okay. Well, if that's the conflict (and it's a legitimate conflict), do we think optimizing short-term or long-term would have value?" Okay. Yes, so let's think that, yes, that does. Our initial thesis, hypothesis was that it would.

Then we took a step back and said, "Well, is there a methodology that would allow you to measure?" This is a math problem. This is not a cold fusion, go to Mars problem. This is just simple math. 

It didn't exist. There was no methodology in the world today that I'm aware of – and I've asked this question of 300 CMOs, so I'm pretty sure we're right on that – that would allow you to assess the value of brand relative to performance. 

In essence, what you're really trying to figure out is what is the value of brand in the way the CFO is going to care about. Nobody knew. 

Oh, my God. I was like, "This is so—" And this is the underlying basis of the whole damn business at some level. 

So, we built a methodology. We raised $2.5 million, and we are now in field with 4 brands to try to assess that. 

We've released two studies so far. Actually, one is not out. 

• We released the study with Ally Financial. Andrea Brimmer did that. She's phenomenal, amazing, amazing.
• We also have another study with Kroger, which the boards have seen but we've not taken public. 

What we know today is that if you run a brand campaign, a brand ad campaign today and it delivers a dollar in revenue (which is a very measurable proposition – that's not hard), we now know through the research methodology we created and figured out and now running that for Ally that was adding a $1.5 in revenue over the next nine months. For Kroger, it looks like the number is closer to $5 additional. 

Now I can go to the CFO (to her or him) and say, "Hey, here's the value of investing in brand in the next year. Can we do that or do we need to have short-term programs right now because there are quarterly goals or some other dynamic?"

It's now a strategy discussion rather than this sort of esoteric, like, "Hey, brand matters long term." It's a very big deal. That's the kind of problems we are trying to solve.

Now, we're going to bring that same kind of level of insight and knowledge rigor to AI and really help marketers to advance themselves here. That's the idea of the MMA.

Michael Krigsman: Please subscribe to our newsletter and subscribe to our YouTube channel. Check out CXOTalk.com.

Aligning metrics, measures, and digital marketing attribution to business goals

The fundamental question then is how to align activities such as metrics, measurements, attribution, or AI back to the core mission of the organization and expectations of the CMO.

Greg Stuart: I think there are two things. One is, what are we going to do? What are the opportunities? What are the use cases for marketing? 

I now have 45 clear use cases that we put together, and I literally will be taking back to the marketer members this coming week, this next week. 

It's Friday today. I've got a meeting on Tuesday and then another meeting Thursday with a working group to say, "Here are the use cases we see that you've told us, we've developed, turned them into hypothesis. Where are these that we're going to focus on? We're going to go do that." That's the primary goal. 

We're also doing a big initiative around responsible AI, which I don't want to lose that. I'm sure we'll come back to that here at some point.

Yes, we're developing initiatives on testable theses and hypotheses. If I can, Michael, let me give you the one that we've done now, the biggest one. 

This is why you can hear my voice, like, I'm excited. Listen. I'm a marketer. I care about marketing. I'd like to see my business saved. 

I think there's way too much [explicit used] in our business we just don't know about. There's way too much stuff we just make up, and it's kind of out of control. And it's part of the reason why I don't think we're respected within the C-suite of organizations. MMA is going to fix that.

Here is the AI example. We had a thesis about a year ago. In fact, it was brought to me by a very good friend of mine named Rex Briggs, one of the most brilliant people in all of the marketing business – no question about it.

Rex said he had been doing some work for vaccine adoption. What he realized is that you can't just put out a broad brand message, a call, because the people who are vaccinated, they're going to get vaccinated, already did. 

He says, "I think you have to personalize this. The problem with health is that you can't use data to do that. You have to operate. You have to do it off of signals, contextual signal data." Not actual data about the person, not personally identifiable information. 

And so, he had run some experiments, and he found that he was able to boost marketing performance – the ability to get people to get vaccinated – by +30% to +60%. In fact, his research showed that he'd saved 3,500 lives by getting that right. Now, that's serious business.

We did what the MMA does. We formed a consortium of a bunch of marketers. GM was in that; Kroger was in that; Monday.com with iHeart Media; and my friends over at ADT, the security company. 

They agreed to participate in a consortium trying to understand what is AI's application to personalize ads and then to use contextual signals to create that personalization. What would be the boost in performance?

Rex has done awesome. By the way, just for those of you with a technical background, we were using one-hot encoding with k-modes clustering, an unsupervised learning. That was the basic dynamics that we used in doing the research – just for those who care about that kind of thing. This is a unique way of bringing marketing to the world. 

We didn't do generative AI and then creative. We put containers around it. Marketers are not willing to do generative AI out live yet. There's no way.

We found a boost in performance +195%. We could 3x the performance of their marketing against the goals the brand set relative to their best efforts in other places to boost performance. That's game-changing. You're talking 2x to 3x (and in one case we saw almost 4x) boost in performance of the ads. That's insane.

To me, this is a revolution. The thing to not forget about, too, Michael, by the way, is that we've also started serving personalized ads that have relevance to consumers. We stop annoying them with ads that don't matter, with [explicit used] that they don't care about. 

That's what we're doing. We're personalizing ads in a way that they might find more helpful, less annoying, less off-target.

To me, the AI thing is perfect. It's the perfect world. Everybody is going to win. That doesn't often happen.

Let's not forget some of this is reducing the cost, too. Tell me one of the things in your organization you have that can lower the cost, that comes in at a lower cost and boosts performance. Those ideas come around maybe once in a decade. 

[Laughter] I don't know. I don't know what the timetable of that is, but it's not very often. It's a big deal.

Michael Krigsman: There are folks who are making sophisticated use of AI and data in order to drive very meaningful results back to their organization that ultimately can be attributed, to which revenue can be attributed.

Greg Stuart: Out of the five studies we've done – I won't name the marketer on this one – one of them didn't work. It was zero. 

Michael Krigsman: Zero. Zero meaning what exactly? [Laughter] 

Greg Stuart: It means that the AI-driven personalization approach that we had did not improve the performance of the advertising at all versus what they were already doing. 

Listen. For me, there's as much learning in the failures as there is in the successes. What the issue was there, we just didn't personalize the ads enough. 

What we didn't know going into this... This is the issue with experiments, right? Those of you in the audience are always focused on having to do that kind of thing. 

Can we isolate the factor? Do we really know what drove it? How do we separate it from what might just be casual observance? 

What we found on this one here is that they didn't personalize the ads enough. As a result, not personalize the ads. 

Now we've isolated it to go, "Okay. It's not that we're just doing better placement of the ads. It's that we're actually personalizing the ads," we think is the issue. We've still got to do more work.

Michael, nothing is simple. Everything pops up other questions.

Here's the issue that came up, too, which was funny. You have to be kind of a real bigger marketer probably to appreciate this.

The issue, too, that we ran into in the first study, we said, "Well, okay. This is great, but how do you write a creative brief when we don't have a finalized ad, when we're just accumulating assets that could be configured into a variety of different ads?"

There are questions like that that come up, too, that we've still got to go solve. There's a lot more to be done here. 

I've been doing this over three decades. I have never seen that kind of boost in performance let alone a lowering of cost. Crazy. It's crazy. 

Michael Krigsman: I don't want to go too much into the weeds, but in this instance where it did not work—

Greg Stuart: Yeah.

Michael Krigsman: –was there a data problem? You said that there was insufficient personalization. 

Greg Stuart: Yeah.

Michael Krigsman: Was that a result of the technology or the data wasn't right or we didn't understand the consumers? What was the problem?

Greg Stuart: It was the color of the product. They had rotated the color of the product, but it wasn't real personalization.

If you're going to do personalization, you want to change the headline. You want to change the background colors. You can do that. You can certainly change the product. You want to show different views of the product. You want to bring in different characters. 

You should even start to assess a reflection of change, slight tweaks of your strategy. I'm not saying you need to be way off your brand strategy, but there should be much more substantive kind of issues.

The changes, the personalization changes just weren't significant enough. The consumer said, like, "Who cares?" 

The point of that is that when we saw limited personalization but different versions of the ads, it wasn't just the placement that then did that, which is what we could have said. We could have said, "Hey," because we're using contextual signal. 

We found another one, too. Michael, can I give you another example of one that we had here?

We did an audio one. This was really interesting. Again, I'll get slightly into the weeds for marketers, but I'll make it fast. 

Most marketers would do a thing called A/B test split on creatives. In this case, it was an audio ad. 

Let's keep it simple. I have a female voice and a male voice. You would split test that, evaluate that, and you would come back and say, "Oh, the female voice did 20% better," and you'd run the female voice everywhere. That's what you'd do. You'd go with a female voice.

We said, "Let's not do that. Let's run male/female, change the message just slightly, but run male/female and see where it went."

So, the female performed better overall – just as we knew. However, male performed better in some states. In fact, it was southern states and Utah and Michigan (oddly enough).

Our thesis – we don't know; we'll prove it at some point – is that those tend to have cultures or populations that might be a little more patriarchal – or guess. We don't know that for a fact, but that was our initial thesis.

The other thing we found, too, is that the male voice always performed better after 10 o'clock at night. The iPhone, the Android, different, so we started to see all of these nuances that we never could have developed enough creative to accommodate. I mean you just couldn't do it. 

You certainly probably couldn't write enough rules to figure that out. You really needed to let the AI itself run and find that best opportunity. 

There's still a lot more work. In fact, we've done some involvement. But that audio one, that audio campaign, really was mind-blowing for us. Again, we're just beginning a new journey here.

Michael Krigsman: It sounds like there is extensive use of AI and data that marketers are using, but it's embedded in various tools to accomplish specific tactical goals. However, some of those tactical goals ultimately roll up to make a major strategic impact. 

Greg Stuart: Michael, give me one more rev on that question. I think there's something interesting in what you're saying there. Why do you ask that question? I want to go after that one a little bit.

Michael Krigsman: Because we began this discussion saying that marketers are not really using AI all that much yet and they're scared.

Greg Stuart: Mm-hmm. Mm-hmm.

Michael Krigsman: But as you talk about it, as you drill in, it becomes clearer that there are at least major experiments that are going on.

Greg Stuart: Yeah.

Driving strategic change for the CMO role and enterprise marketing

Michael Krigsman: But it has to do with things like personalizing ads and placement of ads, which we may not think of as being strategic to a marketer but the end result is this tactical activity will become strategic because it affects directly the marketer's reason for existence.

Greg Stuart: Michael, I kind of sense you were going after the big one there. Jesus. Okay. 

Listen, Michael. We're going to need episodes 838, 839, and 840 to cover this one, just to be aware.

Michael Krigsman: Okay. Well, let's just tell the audience we're going to go for eight hours now. [Laughter]

Greg Stuart: [Laughter] Exactly. Everybody, listen, I'd encourage you to get a snack. Okay?

Michael Krigsman: Settle in.

Greg Stuart: Maybe a box of Triscuits like I have here. Whatever it takes. But we're in it for the duration here. We're not letting anybody go. 

This is how screwy marketing is. Again, I want to be careful here at some level. I'm not trying to pick on marketers. 

I am a marketer. That's all I have ever done in my career. By the way, it's all I ever wanted to do. 

I don't know. I watched Bewitched too much as a child. I don't know what happened. I only ever cared about marketing.

The challenge that we have... When we started to do our marketing org work about six years ago, we had a meeting. It was a closed-door meeting of just marketers, so they would speak freely, and we were just kicking off a project around the marketing org work. The professor, Dr. Omar Rodriguez from Emory University, was with me. 

Wait. Let me set the stage. In the room was the CMO of General Motors, CMO of Dunkin' Brands, the CMO of T-Mobile, the CMO of Allstate, the CMO of Chobani. You get the idea that there were heavy-hitter marketers in the room. [Laughter] Generous of them to spend the time with us.

We asked them a throwaway question. Supposed to be ten minutes, just a warm-up question just to get them going. We asked them, "What's the role of marketing and what's the role of the CMO?"

One hour later, Michael, we never got to another topic. We realized that nobody in the room agreed on the role of marketing. 

Six years ago, we couldn't agree amongst the biggest C-suite executives in marketing in the world on what is the role of marketing. That's nuts. How do we expect the CFO, the board, of the CEO to understand us, appreciate what we do, and be able to give direction to us if we're not clear? –Part of the reason we've done the marketing org work that we have. 

Listen. There's a bunch of foundational issues that we still need to do. 

Now, they answered the question as "We should be about growth." So, the example I gave and the example around this AI research – by the way, if people want to look it up – is called the Consortium for AI Personalization. It's a project within MMA Global that we've now been running and starting to publicize and talk about publicly.

But if it's about growth and you can take the dollars that the CFO or others have given you to invest for the brand – that's how we think, I think, marketers think about this – and we can double, triple, almost quadruple—like I said, we just started. Maybe it's even more—the performance of marketing, I think that becomes pretty strategic. 

Where that's interesting, too, is let's understand the boost in performance of your marketing or advertising is only beneficial until everybody else figures it out, too. 

My point would be... By the way, how many marketers did I say, Michael, have done this? Four, right? Where the F are the others? Why have not 20, 30, 40, 50 marketers?

Listen. I have no axe to grind. I have nothing to sell. I'm not going to sell you another CAP study (Consortium for AI Personalization study). I do it once. I do experiments. That's what I do.

I think everybody needs to jump on this as fast as possible. That's your obligation. 

I don't know. I'm not sure it's not strategic at some level.

Michael Krigsman: On this exact point, we have a very interesting comment from LinkedIn. This is from Deep Khandelwal who says the following. I think he's getting right to the heart of the matter here. He says, "AI comes with a very high cost. Everyone is focused and heavily investing in AI. But what about ROI? Are we able to achieve real business value and growth?" 

Greg Stuart: Yeah.

Michael Krigsman: Maybe you can transplant that question into the CMO realm because it seems right to the heart of the matter.

Greg Stuart: We didn't do anything special for this research here. We used a company called ArtsAI. It's now owned by Claritas to do the actual work here. They were part of the engine.

I'm pretty sure most of those campaigns ran through... Actually, Innovid did one of them, too, by the way. We ran those campaigns through Google's DV360 and Trade Desk. You know there wasn't anything special.

I think where his question may have been going was around LLMs. It felt like the way he asked that. And how you accumulate it to your own. I don't know.

At any rate, LLMs are really expensive. [Laughter] There's no question about that. MMA is trying to build one now for our stuff.

I don't know how far that goes. I've not seen enough of that work to know for a fact, and I'm not sure if we know the real gain on that. 

I think my point, if you take the 45-some-odd use cases that we've built... I'll tell you another one that we're working on right now. We're talking to the guys at MediaMonks, S4Capital, the thing that Sir Martin Sorrell is redoing his agency like.

We're talking to them about a project where we can... Some of you have probably heard this. It's basically you're using agents to summarize your research. 

Let me explain that better. Instead of having a focus group, which takes time to set up and it's a pain in the [explicit used] to do. It costs a lot of money. You've got to get people together, and then there's sort of a falsehood even with focus groups. Focus groups are a little questionable. Sometimes the research there isn't as clean as you'd want to get in the environment that is.

We believe that you can do all that insight research just right off a data set, maybe even just off the Web. And so, we're developing an experiment where we can figure out that.

Now, that would be amazing. If you could take your insights development and shrink that from six weeks to six minutes – not really, but six hours or whatever it is, six days even – you've really benefited because now you're faster to respond to consumers. Faster to respond. The culture, the marketplace is just so much better. 

I don't know. I'm not seeing AI. I guess I have to hear more from them about where they think it's really expensive. I don't know. It may be a competitive advantage to have, so there it is.

Responsible and ethical AI for CMOs and marketing organizations

Michael Krigsman: Let's talk about responsible AI and the ethical, privacy, brand, safety issues. What are you hearing from CMOs about that? –And pretty quickly. 

Greg Stuart: The panel I did yesterday started with that sense of fear, which I think is a very corporate reaction that I appreciate. We do have an expert now who is working with us to develop a series of responsible AI. We'll be setting agendas against that that we need to do.

It's a real issue, and I would also put at the top of that list, too, and it's another experiment we're doing is around bias. How does bias appear? 

Everybody understands. I think your audience all gets it. The data sets are only based on what culture has told it already. That means bias gets repeated. 

We're going to two things. One, how do we do our best to remove bias? It's important to marketing. Also, I want to try to understand the economic value of getting at bias because I think it will give us an extra boost to be even more serious about that topic.

Again, more to come around that. But yeah. I don't know. 

We've got a list of about 25 things that we're running by the team. We formed this AI leadership coalition that's going to be trying to help dig in to figure out where we need to be better around guideline frameworks. 

I don't know. I could go on and on. But yes, critical, a critical topic. You're right, Michael.

Michael Krigsman: The CMOs that you're in touch with obviously are recognizing the crucial importance of these issues as well, I'm assuming.

Greg Stuart: Of course, and boards' jobs are to manage risk, so the CMOs are getting asked by the boards, "What are we doing to mitigate risk?"

The tone that I heard in this discussion yesterday that everybody was really focused. The disappointment about that conversation is they're focused on risk first rather than the huge opportunity. 

I'm aware of the opportunity and not everybody is. We just need to two. Everybody has got to have two workstreams: performance and be responsible. 

Yeah, let's go. [Laughter] It's not that hard, everybody. Let's go.

Michael Krigsman: Among CMOs, do you think that there is really broad and deep understanding of what the opportunities of AI are or is there a lack of understanding around what's possible?

Greg Stuart: Anything new is hard. I think companies, corporations in particular, struggle to get to new. It just is what it is. 

I was there for the dawn of the Web. Not the Internet but of the Web in '94. I was running an interactive group at Young & Rubicam at the time. 

I mean they're just slow. I don't think we're as slow as we used to be, but we're not as fast as the opportunity presents itself. How is that? 

I don't know. I don't have a better answer for that, and I can't manage people's individual careers or the risks that they have. And they've got other complexities within companies. 

But I don't know. We see big... I don't know. I think I said enough. It's a big opportunity. Let's go.

I think most people do think that marketing is probably the area that will be most impacted by AI, though. I will say that. I hear that a lot from people. It's generally what we seem to think.

Impact of AI on marketing teams, talent, and staffing

Michael Krigsman: What do you marketers think the impact of AI and the shift in skills will mean for their own organizations?

Greg Stuart: I can't speak to everybody on what they should do around that. This research I mentioned before around marketing org that we've done and that we're trying to adapt around AI is really around capability and, in some degree, competency building. 

I think the thing I would say – and this is a little snarky – is you've heard people say that AI will replace jobs. I think the issue is actually different than that. I think the issue is that if you don't learn AI and how to apply it to your daily job, you will get replaced by somebody who does know how to do that.

That's not the machine. That's your lack of adapting into the future. 

It's certainly what I try to politely communicate to the team of the MMA that we have a responsibility as forward leaners in the industry to really get good at this stuff as fast as we can, and we're doing a bunch of programs around that. 

Yeah. That requires the energy and effort of the people at the top to really direct the teams to go there. There's a bunch of things that we do around that, but that's a whole other topic, I think, at some level, but critical – critical, critical, critical.

Michael Krigsman: Essentially, I'll paraphrase here. Tell me if this is wrong or right. the bottom line is AI is going to have an enormous impact, and you better learn about it, and you better get your team and your organization up to speed. You have some time, but you better be on top of it.

Greg Stuart: Yeah, yeah. Come on, people. Abacuses are dead for a reason.

Michael Krigsman: [Laughter]

Greg Stuart: Let's go. Game on. The ball is in play. Make it happen. Get there fast. If you don't do it, we find somebody who will.

I don't know. I don't know. I've always been very future-leaning myself personally for whatever reason. So, I'm sensitive to those who maybe don't feel that way. [Laughter] But really, you've got to get going.

Michael Krigsman: I love that, "Abacuses are dead for a reason."

Greg Stuart: [Laughter] Right.

Michael Krigsman: Okay. That works. 

Greg Stuart: I don't know if it was that clever, but yeah. Fair enough.

Managing the risks of AI, including brand and reputation risk

Michael Krigsman: Another question, from Twitter this time, and this is from Lisbeth Shaw. It's a good question. She says, "How should CMOs manage the risks of using AI, including brand and reputational risk?"

Greg Stuart: We're going to try to create (using crowdsourcing of our members) to the best of our ability (and the experts that I have at our disposal) a series of guidelines, checklists, that will mitigate some of that risk. 

I think, too, you want to be careful. You want to manage, obviously, management around the risk. You want to have buy-in from the CEO, CFO, whoever you report into. Structure, too, is a pretty big one around this. There are also places, too, where it really matters. 

Rex Briggs, who I mentioned earlier, just completed a book. It's not out, public yet, but I've had a chance to read it. It's called The AI Conundrum. 

Rex, he's spoke at our event, so if you want to see more, I can get a hold of that for you. He talks a lot about understanding where you really have to be exactly right and other places that you don't. 

There's a framework that he has in the book that helps you think those through and your risk tolerance should be lower in those cases. It's matching that up. 

Listen. Automated driving, self-driving for cars, we'd better be more right than wrong. And we'd better be really damn good about it because people die.

Marketing, eh, okay. There's some risk, but not catastrophic often. 

We'll do some more work on that. Good question, though.

Michael Krigsman: Deep Khandelwal comes back, and he just wants to comment that you're a great storyteller.

Greg Stuart: [Laughter] You know, and that, I'll give you a free... I'll give you Rex's book for free. How is that? If you write me, I'll send it to you. [Laughter] 

Michael Krigsman: All right. 

Greg Stuart: I don't know if you're right, but you're kind to say. [Laughter]

Michael Krigsman: Greg, as we finish up, final thoughts on marketing today and where you see CMOs need to be focusing their efforts.

Greg Stuart: You all watch the patterns, right? It was the PC. Then it was the Internet. Then it was mobile. AI is the next trend. This is the most important thing that's going on, and this is where competitive advantage gets created, where opportunities anew are accelerated. 

I've built my whole career on following all those trends all through the decades. So, I would strongly encourage you to lean in as far and as fast as you can. 

As I mentioned earlier, in the research we've done around performance of marketing and advertising itself, I've never seen those kinds of gains. And we've just started. The tools aren't even really developed somewhat.

I'm a strong believer that if you really want to move your business forward, if you really want to play a role, if you really want to move to the level that we want, that MMA wants CMOs to get at, this is that opportunity here and now. Let's go, everybody.

Michael Krigsman: Well, it's good advice. With that, I want to say a huge thank you to Greg Stuart. He's the CEO of MMA Global. Greg, thank you so much for being here. I really, really do appreciate it.

Greg Stuart: Thank you, Michael. It was really an honor to get an opportunity to join here at CXOTalk. Thank you.

Michael Krigsman: Everybody watching, thank you for joining us, especially to those folks who ask such great questions. Now, before you go, please subscribe to our newsletter, and subscribe to our YouTube channel. 

Check out CXOTalk.com. We have extraordinary shows coming up. Next week we have the CEO of Salesforce AI. She's amazing, and that will be great. So, check out CXOTalk.com. 

We'll see you again next time, everybody. Take care. Bye-bye.