Personalized advertising: a boon or a bane in the AI era?

In this episode of the Greenbook Podcast, we are joined by Mark Truss, Chief Research Officer at VML, who shares his journey from vendor research to advertising agency work, highlighting the distinct methodologies of each sector. Mark discusses the evolving landscape of analytics and research, emphasizing the role of AI and automation in enhancing research efficiency and creativity in advertising. He explores the challenges of personalization in advertising, advocating for an integrated approach to brand and performance marketing. Additionally, Mark speaks about the importance of sample quality, his role as a part-time adjunct professor at Columbia, and even his involvement in a musical group "The Outliers".

You can reach out to Mark on LinkedIn.

Many thanks to Mark for being our guest. Thanks also to our producer, Natalie Pusch; and our editor, Big Bad Audio.

Hello, everybody. It’s Lenny Murphy with another edition of the Greenbook Podcast. Thank you so much for taking time out of your day to spend it with myself and my guest. And my guest today—we’ve been trying to get this scheduled for over a year, and the fates just did not align, but now they are. And I’m so excited to have Mark Truss, the Chief Research Officer of VML on. Mark, welcome.

Lenny, thank you for having me. Glad we finally got this worked out.

Well, you know, I’ve been bugging you for—

[laugh].

I mean how long have I been, “Mark, come on, you got to come on.” So…

I’m glad it worked.

I’m glad it did, too. So, Mark, for those who don’t know you, why don’t you give a little bit of your background and your CV, if you will, in a condensed form to kind of set the stage?

Sure. So, I’ve been in the industry just shy of forever. I spent about the first half on the vendor research side of the business at the NPDs and Ipsos’ of the world, then came over to the advertising agency side of the business. I’ve had one company, three different brand name business cards because of mergers and acquisitions, VML being the most recent. I’m a part-time adjunct up at Columbia teaching research and insights and strategy as part of a graduate program. I sit on a couple of boards. I’ll put that out there now because I may mention them somewhere along, and just for transparency. So, I sit on the ARF’s Board and I sit on the AEF’s Board, the ANA’s Educational Foundation. And maybe, as you can tell, I’m a musician, so I play in a band of research folk, actually. So, of course, we are called the Outliers. Right? Because it had to be something research—

Oh, so, with Joel Rubinson? You play with Joel?

That is correct. Joel plays with me. You got that right, Lenny [laugh].

[laugh] All right.

Exactly right.

Very cool. I didn’t make that connection. I did not know that you did that. And I assume you’re—based on the fender, you guys—audience, you’re not seeing this. Mark’s wearing a fender hat, so you play guitar?

Correct.

Okay. All right. While Joel gets up there with his harmonica.

Exactly.

Okay [laugh]. Very cool. Very, very cool. So, first question, you know, knowing that background, what was the most surprising thing going from the—coming from the dark side. Right?

Yeah.

The supplier side, into the agency? Just kind of frame that up. I think the context of now, your perspective is, you know, this now working through the agency side.

Yeah. I mean, it was a bit of culture shock, for sure. And I think it’s more about the orientation of a researcher in those two different types of organizations. So, at a research vendor, it’s much more about getting to an answer. Right? So, a cone with a big part at the top and a small part at the bottom and really trying to drill down and get to an answer. But at an agency, it’s almost flipped the other way. If the vendor side is more convergent, let’s converge on an answer. Agencies often are the other way; they’re divergent, which is: here’s some ideas. Here’s a problem. Here’s some hypotheses. What are all the different ways this could come to life? What are all the different dimensions and contextualizations of a consumer experience as it relates to that topic that might be pertinent to what we’re trying to do? So, it’s much more exploratory, might sound trite, much more creative in the way you approach it. It’s not so what is the right answer? It’s what’s the realm of possible right answers, and what could those be? So, to me, it’s a very different approach. I mean, we’d still do some of the stuff you’d see at a vendor, which is, you know, which of these ads performs best and why. But it has much more of an exploratory feel to it often.

And how’s that played out over the—we’ll avoid the dreaded two letter word for the moment.

[laugh].

We’ll get [laugh] into that. But just the advent of digital overall and the focus on analytics and information that is simply available by default because of the form factor of the marketing. So, what kind of—what does that evolution look like? And then maybe we’ll spring off of that for that other topic [laugh].

Yeah. You know, in the beginning, I think, when—I don’t want to say when analytics was born, but I want to say when it became sexy, which were two very different things. Right? Analytics have been around for a very, very long time, but probably late 2000s, it sort of became the sexy thing due to digital. I think at first there was some competition between analytics groups and research groups until we figured out just how adjunctive and complementary the two types of research and insight are and how both can be used to great quantitative effect and great qualitative effect. But there was—we clearly worked out pretty quickly that there was sort of a cadence. Hey… we found this. What could you do to help us understand why or what? Or you found that. What can we see that’s actually happening in the digital world that supports that? Or how do we bring the two data sets together? Right? Sort of what I used to call research alchemy. And it took a little bit, but I think we figured it out. And, you know, we’re sort of, I think today, in a world where it’s more about specialties than it is about department names. And on almost every project assignment, whatever, there is someone from the more traditional research world, someone from the more traditional analytics world working together and kind of figuring out what’s the best way to do this.

What would that flow look like? I mean, so let’s say, okay, there’s a creative that we want to test because we have the ability to do traditional testing. We’re also live A/B testing now. So, can you give, you know, kind of the audience an example of what a—you know, I’m sure nothing is typical, but I’m also sure that there is an approach, a flow, that, you know, this works most of the time. And what does that look like?

I think it depends where you are in the process. Right? So, at an advertising agency, there’s a process that begins with, you know, collecting data, understanding all the different dynamics in a category. What do the—does the target—who is the target audience? What lights them up? How do we segment them? All that kind of basic blocking and tackling, that is probably the vast majority of the time before you even get to a situation where you’re creating ad-like objects. So, sort of discovery, strategy, understanding up front, and then, you know, as you develop ad-like objects, that becomes more about testing in some way, shape, or form. Maybe it’s an A/B test; maybe it’s a piece of research, and then measurement on the very back end. You know, how did it do? And I think, at the beginning of the process, it’s probably in equal parts analytics and research working side-by-side, as I described before. As we get more into testing, it often becomes more analytics now in the digital world. Right? Although sometimes, it’s research. And I think a lot of the question on are we doing research to understand what’s going to work in market or effectiveness or is it analytics is often about what else do we need to know. If we just need to know what’s performing better, then analytics is a really good approach to do that. But, if we need to understand why, and we need to understand specific segments of people, like, you know, how close they are to making a purchase or something like that, then we need to include some research functionality in there, too, to ask those questions that we’re not necessarily going to be able to get in a strict A/B or an A/B that has a five-question component to it or something like that. So, I think it really depends on what we’re trying to do and where we are in that process.

And what about the incorporation of—since you mentioned why, right, I automatically go to the advent of kind of non-conscious measurement in all of its various and sundry forms, right, from implicit test to, you know, facial coding and everything in between as a key input, particularly around creative. Is that accurate? Was that—have you seen that toolkit become more important as input to this process? Or do you think that we’re able to capture that through other data sources equally as well?

I guess, when I think about implicit and biometrics, I kind of think of them in two different classes. So, one might be the why, but the other might be the reaction that you’re looking for. That’s somewhat hard to get in some methods. I’ve always been stunned. You know, I’ve been a big fan of biometrics for 20 years, since they’ve been talking about it. And every year the GRIT Report comes out. I always take a look at where is biometrics on the adoption curve. And I’m always a little dismayed that it remains niche, shall I call it? Right? I don’t know what it is now—10 percent, 15 percent, something like that—hasn’t necessarily become mainstream yet. And, you know, in different—well, I’ve talked to Kantar over the years about their experience kind of including biometric components in their traditional copy tests. And what they always say is, you know, “It’s helpful in very specific applications, but it’s often adjunctive. It’s not a replacement for what we’re doing.” So, I love it because I’m a firm believer that often we need to look below the surface to really understand what people are saying. People often—and it’s not because in research people are being—they’re trying to lie or present a point of view that’s not really what reflects their true motivations. It’s that they often don’t have access to them. They don’t really know. Most people, you and me included probably, don’t really know what motivates us to do certain things, or we misattribute our motivation to something that sounds plausible, reasonable, or makes us feel good about the decisions we make as opposed to what really motivates us. So, trying to get to those types of things, to me, are really important. And that’s where I think biometrics play a big role.

Yeah. Agreed on all points. And I would argue that it’s becoming more important because of the fragmentation across media channels. Right? I mean, you know, pinpoint targeting for very specific populations. I mean, I know my media habits are radically different than they were even just two years ago. So, it’s very specific, very targeted. And I’ve observed that the advertisers are not targeting me very well with some of the ads that I see on some of the channels [laugh] that I absorb, you know, media on. And I really just need to reach out to these folks and say, “You got to do a little bit better to understand what motivates folks like me and my cohort that are engaging here.” So, do you think that’s—is there accuracy to that?

I do, and I think this is part of a much bigger question about performance versus brand. Right? And that discussion that always kind of dominates industry conferences and panels on [benean field], IPA, types of long and short of it. You know, is it performance? Is it brand? And I think the industry has gotten off the rails a little bit, and it has fallen a bit in love with performance as a tool. And I understand why. A lot of pressure on marketers to perform and much easier to measure performance than it is brand marketing. But I think the question is completely wrong. Right? It shouldn’t be about is it performance; is it brand? It’s about what’s the mix and the degree to which both. And the mix is both in terms of a campaign. Right? Which is the way benean field often look at it, that 80 percent of the marketing should be performance versus 20 percent, depending on category and maturity and different variables. But also within a campaign itself—you know, I don’t know if you know Andy Smith, who runs insights at Flowers Foods, but—

Oh, yeah.

—a conversation I was having with him once, he said something really—I thought it was really smart. And he said, “You know, you can talk as much as you want about performance and brand marketing, but I’ve never seen a campaign that didn’t carry an expectation that sales were going to move. I don’t care how ‘brand’ you thought that was, how much you’re investing for the future, there’s still an expectation that it’s going to move product.” Right? So, you know, when I heard said, “You know, you’re right. Everything is performance marketing. It’s shades of where the focus is on those.” But, I don’t know, maybe I didn’t answer the question. That’s what it made me think about when you said that.

No, that was great. And, yeah, Andy, shout out to Andy. Andy’s great. And, you know, I’ve always framed up in my head that kind of the marketing lifecycle, just overly simplistic, is to engage, understand and activate. Right? That’s the process. And that activation, that’s that performance piece. Because, ultimately, our job is to sell stuff. Period. Right? Actually, I had somebody, an executive at Nielsen years ago, they were trying to recruit me. And we—anyway, that’s a long story. But asked me what is the ultimate goal of insights? To answer questions. No, the ultimate goal of insights is to sell more stuff. And whatever that stuff is, a concept, a product, it doesn’t matter. And that we often, I think, to your point, we lose sight that that’s—there’s this funnel, this process we have to look through, and everything has to serve towards a step of getting to drive to sell more stuff.

Ultimately, we’re selling something, right? Whether that’s an idea, whether that’s a point of view, or whether it’s a real product that people have to put coin down for. You know, coming out of the brand world, I do understand the role of brand in propping up that sale at the end. And I think that’s what benean field often talk about is that it’s not brand or performance. It’s not brand for the sake of building a brand. It’s that, when you build a brand, it makes the performance side work that much more powerfully and that much more easily to make the sale. Right? Because people already know what it is, they know what it means to them. They have a certain feeling when they experience it, all that good stuff. Right? So, again, it’s the wrong question to be asking. Brand or performance. Brand or performance. It’s really about how do you integrate the two of them in a very smart and clever way, such that you’re making all of your media investment work harder for you.

I love that. That’s fantastic. Now, that’s probably a segue then into the topic du jour. So…

What’s that? What’s that, Lenny? I don’t know [laugh].

[laugh] So, how has the last year been for you? What do you see happening on the impact of AI within—from an advertising or publisher, marketing, whatever context you want. And then we’ll get into, you know, what do you think it looks like in the future. But what’s your—what’s last year been like for you?

I think it’s very early days, very much about experimentation. And, you know, early on this year, you know, I sat down with my team and said, “This is going to be a fun year, just watching all of the different ideas and applications people come up with for how to use AI. Right? And it’s going to be fascinating, and we should be a part of that. Right? We should be out there playing around, seeing what it can do.” You know, and early on, we learned some really fun stuff. You know, we were doing a survey on—I can’t remember the topic—and I just went into, you know, ChatGPT at the time and just asked it, “Can you write me a survey about this, this, and this? Here’s the background. Here’s what we’re trying to understand.” And it did, and it was awful because it’s never been trained on how to write surveys. Right? But it came up with, like, five or six questions that we hadn’t even considered to put in the survey. And I said, “Oh, this is great.” You know? And this was a case where I said, “This is really good because it doesn’t have to be right. It just has to be interesting, and it has to be inspiring, and it has to take us—our thinking in a different direction.” Right? So, we said to the team, “Great. Check the box. This is a really good use for this kind of stuff.” We played around like everybody else with could it code unstructured data for you, and how well does it do it? And there was something on the [wonx] list at that time—someone had put out there. I can’t remember who it was. Sort of the query to put in. Right? They had worked out a query that was quite detailed. So, I tried it, and learned, oh, this doesn’t work. That does work. Oh, it doesn’t understand that each case can have multiple ideas. I need to tell it that that’s okay. Right? So, all these things—but great time saving things. Right? So, I think, you know, people are still trying to figure it out and figure out where it can bring value and where there are ‘watch outs’. I don’t know. I look at it very much like a really quick, smart intern that you can deploy on certain things, and they can do it way quicker than you can. And it’s not 100 percent. It’s not perfectly accurate, but, again, I’m using broad-based ChatGPTs and Perplexity AIs that aren’t really trained for this purpose. I’m assuming, once those tools are trained specifically for these applications, they’re going to get much, much, much better. I haven’t seen it yet. You know, it’s funny because the—one thing we’ve noticed, as we talk about this, is now every research vender in the world has an AI solution, which kind of blew me away. I’m like, “What were they all doing last year?” You know, now, suddenly, they all have it.

[laugh].

And I was talking to Forrester recently—because we subscribe to their service—and they asked us, you know, “What could we do for you? What are some gaps in your needs?” And I said, “Here’s what I need. I need Forrester, you know, to come out with, like, an AI BS detector that can be deployed against all of these people claiming to have AI,” which isn’t AI, well, in many cases, and just really say, “Hey… these people are really doing it. These people aren’t.” Because, right now, everybody says they’re doing it, and much of it is not really AI. They’re just algorithms being called AI. Right? But I’m really bullish on it. I mean, I think it’s, like, the greatest thing that’s going to help our industry in many, many ways. But we don’t even know what they are yet. I don’t think we even know what the questions are to ask. You know?

I agree. I’m actually really curious. You know, last year, last wave of GRIT in the spring, we started asking these questions of—we’d always asked about automation. Now, we specifically got into generative AI. Anyway. And we just fielded it again. The cross-tabs are actually sitting in my inbox, and I can’t wait to get in and see what changes. And particularly, I think related to your point, for years we’d ask the question we called, “The day in the life of a researcher.” Right? How they were spending their time. And we started doing that with the era of automation. Right? The [Zapies] of the world, those companies emerging. And thinking, “All right. Now, we’re going to see people that they’re going to spend more time doing, you know, value added, you know, strategic stuff, not tactical stuff.” Never saw that, ever.

Yeah. Yeah.

[laugh].

Yeah.

Nope. People just did more of the same.

Yeah.

That’s what automation allowed. So, I’m curious to see, now, whether—to your point of generative AI, are we just seeing that people’s bandwidth is just increasing even more of doing more of the same, or will this be the tool that allows us to do some different things?

I think the latter.

I think the potential for the latter is there, absolutely. Even the use case of—to your point of just dealing with unstructured data. So, such a huge time savings. Even though it may not be perfect, it’s pretty damn good. I’ve been a fan of text analytics, and it does so much better than anything with text analytics. Now, what about on the creative side? Because, I mean, you know, the stuff that we talk about is all, with AI applications, it’s really kind of geeky and wonky. But some of the creative stuff that I see happening, it’s like, “Oh, my God.”

Yeah.

I mean, really, like, literally saying, “Make me a video showing Mark and Lenny fighting aliens.” Boom. You know? It’s insane. So, what do you see happening on that front? And what are your [laugh]—what do the agency folks think about that, especially your creatives?

Yeah.

Are they like, “Yeah, this is great?” Or like, “Woah. No.”

Not—I haven’t seen much on the production side, which is kind of what you’re talking about. Right? But I do think it is being used on the inspiration side. So, early in the creative development process to get ideas, to play with images in your mind that you may then, you know, turn to another tool to actualize in the real world. And, when I say that, I’m talking about big agencies catering to big clients, so a company like VML who, you know, has Ford as a client, right. For those types of agencies, I don’t think they are using AI as ‘the’ creative tool. Smaller agencies, who are maybe catering to small and medium-sized businesses, or the small and medium-sized businesses themselves, it’s just like a democratization moment. Right? Where suddenly they have access to up their advertising game, where they couldn’t’ really do that before. And there’s a couple of tools out there that I think do a pretty good job of that. So, not seeing a ton right now as a tool. Although, you know where I am seeing it? Where AI is the advertising idea. Right? It’s not that they used it to get to the idea; it is the idea. And I’ll give you an example. And this came out of VML, and it was for our client, Sherwin-Williams. And what they used AI for is instead of going and picking out color chips at your local hardware store or going online and doing that, you could articulate to an AI machine, a chatbot, what you wanted that color to feel like and look like. So, people would say, “Hey… listen. I’m looking for an ocean breeze. I want a color that evokes an ocean breeze on a warm, Caribbean beach,” and it would bring up a color. And then you would say, “You know, a little less blue and a little more…” And it would custom design a color for you based on your language, and then give you the formula for you to bring to your local Sherwin-Williams store and say, “Here. I want this color.” And I’ve seen a few instances of that, where the AI is the creative idea, as opposed to using AI to come up with or produce a creative idea. So, I’m seeing more of that, which I think is really interesting. Right? And, in a creative company, maybe I should expect no less. Right? Me as the research guy [laugh], I would think of it more in a literal aspect, but they are thinking of it in a much broader aspect, which I think is kind of cool.

Well, that—back to the point around targeting. Right? I mean, we’ve been talking about personalization forever. I always think of Mark Pritchard’s quote of, “P&G wants to have a one-to-one relationship with everybody on the planet in real time,” which means the right message to the right person, you know, et cetera, et cetera, which means personalization at scale. And it would seem to me that this technology creates that virtuous cycle of rapid, agile, iteration, personalization through data synthesis, and then built-in, you know, measurement tools as well, so that we’re always optimizing that. And I’m of two minds of that. So, there’s—in this [laugh]—I always use the example. I used to present, and I would use the example of Minority Report, and Tom Cruise walking through the mall and, you know, the personalized ads and thought, “This is our future, and that’s so cool.” I don’t know if I think it’s that cool anymore. Well, I still think it’s cool. I don’t know if it’s good. Right? [laugh] I go back and forth.

Yeah. And it’s got a high, creepy effect.

And it seems much more black mirror to me now than it did before. But it would seem that we’re there. And, again, from an agency standpoint, do you guys see that, potentially prepping for that future of, “Yeah, I think we’re maybe not quite there, but we’re almost there.”

Yeah. We’re definitely there. It can be done. We have done it. And I don’t think we’re unique in being able to do it. I think there’s many people who can do it. If I think about the Gartner Hype Cycle, I actually think personalization is currently, if I remember it right, down in the Trough of Disillusionment section of the curve. And I’ve been involved in a few projects where we’ve done it. And some of the problems with it is more in the messaging, so the right message part, not the right time. Right? The right time part is easier to do. The right message, it’s a little tricky, especially on the client side if they’ve got to figure out regulatory-wise, brand-guideline-wise, how do they keep it all together if we’re all getting different messages, right? So, what’s the danger to the brand if everyone’s getting a different message, right? Now, in some brands, maybe that’s okay. Maybe if it’s Coca-Cola and it’s all about happiness and joy, everything has to stay within the happiness and joy idea. But if there’s a campaign idea—or, you know, that sort of cohesiveness of campaign elements and campaign messages aligning with a brand idea, it becomes really complicated if you’re serving up ten, twenty, thirty, forty, a hundred, a thousand, a million different messages, it becomes really challenging to do from a brand point of view. So, I—you know, I think we’ve played around with it a little bit. We’ve had some success with it, but there are some sort of unforeseen barriers. And, you know, we actually tried it in pharma, which became a pretty big problem because all ads have to go through regulatory review. Right? So, to suddenly go to the regulatory team and instead of saying you have one ad to put in front of you, “Oh, yeah, we’ve got 30 [laugh].” And they kind of look at you like, “What are you doing? Why are you doing this, you know? We can’t handle this.” So, there are some hurdles in the real world to actually making it happen. Or maybe they’re not hurdles. They’re considerations about how one approaches that. But it’s happening in certain places, for sure.

So, then—and being conscious of your time and the time of our listeners, almost final question. Based on everything we just talked about, and I think especially that idea of personalization, what prediction do you have on your seer at—over the course of 2024 that will be most impactful to the insights and analytics industry. What are you just waiting to see drop? They said, okay, this has changed the game or moved the ball significantly down the court in one way or the other for the industry?

I mean, I think AI and automation will. I don’t know when. I don’t know who is going to really do it well. You know, I know there’s plenty of companies out there that are doing real AI. I know Ipsos and Kantar are doing a lot of that, which is interesting. Of course, I want to peek under the hood and see how it’s working. Here’s what I’m hoping for this year, and maybe this is narrow, but it’s a huge pain point for me, and it’s sample quality issues. Terence McCarron, CEO at OpinionRoute, had put out a prediction for this year that—and I’m not sure I’m going to get it 100 percent right—but it was along the lines of in this year, sample provides are going to back away from the implicit assumption of quality, which made me kind of go, “Hey… hey… hey… hey… what are you talking about?” [laugh] You know? I think we already have that problem, and we need to figure out how to fix that. But then, when I think about two trends in sampling, right, one being synthetic sample, and one being custom recruiting, so the [RegDatas] and NewtonXs and all experts of the world recruiting quant like qual used to be done. Maybe what’s going to happen this year is we’re finally going to see different quality tiers at different price points based on this. Right? And I’m all for that because, to me, that’s not transparency. And, you know, I’ve always been of the mindset I’m happy to pay more for better sample, and it’s not because I’m a purist. It’s because it actually saves me money on the backend because I don’t have to have someone spending hours going through data. Right? I think this is a big problem in the industry, and I’m dying to see some movement there. And I’m hoping this is the year. And I’ll give you one little tip that my team stumbled upon in terms of sample quality, quite accidentally, a couple years ago. It really helped us—is we went through the process of pre-matching sample provider panels to identity truth sets. So, an example was matching Prodege’s panel to Experian at the name, address, city, state, ZIP level. Right? And then when we would do research and look and say, “I just want my sample among matched panelists versus a test group of non-matched, the fraudulent rates, or the ones we deemed as fraudulent, dramatically different.” In the matched panel, it was about one percent. In the non-matched, it was about 30 percent. And, to me, that’s, again, another tier of sample quality. These are matched. We know who they are. They’re real people. We could send them a letter at home if we want. Right? So, those kinds of—I think this would go a really long way to helping the industry be clear that we have trusted, quality sample out there.

I could not agree with you more. I’ve actually been working with a—I mean, you know, on my Gen2 side, I do a lots of advising and consulting and all that good stuff. And one of my really fun clients right now that I’m working with is a panel company. And we were actually just working through this last night because I fleshed out—like, look… there’s a graph here that the more strategic the issue and the more impactful the decision that needs to be made is, then quality matters more. The more tactical or exploratory it is, quality matters less, therefore price. And there is a role for low-quality sample. What I’m not worried about—is five or ten percent difference going to change my decision? It’s irrelevant. Save the money. So, if five or ten percent is a multi-million difference, then, hell yeah, I want to make sure that it’s absolutely appropriate. And, of course, you know I love the idea of matching. I think you recall, you know, Veriglif, my attempt to try and—

Yeah.

—link these things and do that. Which, too early. That’s a whole another conversation. But… But I do think that there’s enough momentum around the topic and enough pain being felt by folks like you, and you’re talking about it, that that’s what’s going to move the needle. And I think that’s probably the biggest issue is that buyers are no longer being quiet about what they’re seeing.

Yeah. Which I think is important.

Yes. And we see it impacting, you know, some companies that are public. We see their refund rate really affecting, you know, their evaluation. So, that’s—those are all big drivers. Mark, what’s the—what did you hope that we would talk about that we didn’t get around to? Was there anything that you wanted to make sure that we brought up?

I don’t know. You know, I—AI was the thing that I thought we would sort of really, you know, jump on. I think the other thing that we haven’t talked about that I think is a big deal is—in social, specifically, is influencer marketing and what that means to the industry and what it means to the insight industry. How do you measure it? How do you… I mean, it’s kind of like the spokespeople of old. Right? What does the spokesperson stand for? Do they enhance our image? Are they contrary to our… you know, what are we trying to do there? And all the issues around influence, which I find really interesting—there was a report in JAR, a paper, I think, out of Australia, that found that with influencers, there is reverse double-jeopardy.

Bud Light? Right? Is that an example, for instance [laugh]?

Well, sort of. But what—not that jeopardy. Right? So, the double-jeopardy of small brands tend to suffer in two ways. Not only did—this is [Aaron Burke Bass] kind of stuff. Right? Not only do they have a smaller user base, but they’re less loyal to them. Right? And what they’re finding with influencers is that the larger the influencer, the lower engagement with them, which kind of intuitively makes sense. Right? If you’ve got five million people following you, it’s difficult to get them all engaged. But if you’ve got 10,000, it’s easier. But it’s sort of—I mean, these are things we’re sort of learning in the marketing world as influencers become more important. How do we quantify their value to the brand? How do we know which ones are the right ones? And how do we sidestep Bud Light issues, right, to your point? So, it’s just sort of an interesting area that’s becoming used more and more and more with marketers. But I don’t know that the insights world has done a lot with it, and there’s probably a lot more that could be done there.

I agree. That’s a whole other topic. I would love to talk to you offline about that.

Yeah. Sure.

I’ll say this, and then we’ve got to wrap. I’ve become a huge fan of Substack because of the experts. Right? And there’s such a variety of wonky-ness [laugh], right, of geeky-ness on specific topics on Substack, whatever the topic may be. And they are influencers. There are certain people like, “Oh, I have a question about this. I know this person. This is all the hell they ever write about is this. I trust their opinion.” And that audience is growing. So, there’s a whole other channel, I think, besides TikTok and, you know, and X and all that of the expert influencer emerging in a very fragmented media ecosystem. But, anyway, that’s a whole other conversation.

Yeah. And here, B2B.

Yeah. Absolutely. Well, Quora for B2B, yeah. So, we can circle back around to that. Mark, I’m so glad we finally had this chance to chat. It’s long, long overdue. We’ve circled each other for a long time, and this is great. And I hope it’s the first of many other conversations.

Yeah. I would love that. Thank you for having me.

No, thank you for being here. Where can people find you?

Reach out on LinkedIn, or you can ping me at VML.com, mark.truss@vml.com.

All right. Now, you better be prepared for some—now, you put that out there, the email, so you may—

That’s okay. I like it.

Okay. All right.

[laugh].

Mark, thanks so much. Have a wonderful day. Thank you to our listeners for spending time with us. Big shout-outs for our producer, Natalie; our editor, Big Bad Audio; and to our sponsors and supporters because without you Mark and I would have fun talking, but we wouldn’t be so focused on trying to add value to others. So, that’s it for this edition of the Greenbook Podcast. Everybody have a fantastic day until the next time. Bye-bye.