The Human Pillar

Speaker 1: Welcome to the Business of Marketing. My name's John Horsley, I'm the host of the show today. I have the absolute pleasure and delight of being joined by Jason Mander, who's chief insight officer at GWI. Welcome to the show.

Speaker 2: Thank you so much.

Speaker 1: So you've been at GWI, as you were saying a moment ago, for the last nine years. What was it that attracted you into the world of research in the first instance?

Speaker 2: Yeah, it's a great question. I actually started out as a research fellow at the University of Oxford. Uh, believe it or not, I was studying classical archeology, and went on to complete a PhD in the social sciences, understanding how families across the Roman Empire were behaving. And towards the end of my time there, it became clear that the skills were transferable, so I went from predicting what Romans were doing 2,000 years ago to understanding how consumers across the world now are thinking, acting, and feeling.

Speaker 1: I love that.

Speaker 2: It's quite a jump, I'll grant you-

Speaker 1: Yeah

Speaker 2: ... but it works.

Speaker 1: But I'm wondering, in some ways, if we're come, kind of coming full circle and, and back to the start again.

Speaker 2: Yeah.

Speaker 1: Um, for the listeners, it would be really good if you just explain a little bit about, um, your business and its, its recent detrás, as it were.

Speaker 2: Sure. So GWI has been around since 2009. We were founded to revolutionize how market research was being done, going digital with global reach. Fast-forward to right now, and we're leading the charge in terms of how human research or intelligence needs to be within the AI ecosystems which everyone is using to make decisions. So our data from, uh, all the countries around the world, more than a million individuals per year, is being, uh, fed into all of the AI tools that our industry is using so that any decision can be grounded in real human intelligence. How people actually think, act, feel, and behave.

Speaker 1: Yeah, and can you give me an example of some of those sources? I'd imagine that they're, they're broad and fairly disparate.

Speaker 2: Some of the sources?

Speaker 1: Yeah.

Speaker 2: So, um, really it's all interviews with real-life individuals-

Speaker 1: Yeah

Speaker 2: ... in more than 50 countries. We speak to tens of thousands of people, all carefully researched to be fully representative of the country that they're in. And we ask them a wide range of questions about their lives, their lifestyles, their digital behaviors, which apps they're using, which brands they're interacting with, which things they might be considering buying. And when you have all of that data on a global scale for more than 15 years, the trend ability is amazing, so you can see how behaviors are moving, how attitudes are shifting. You can see adoption rates of new technologies. That's at the absolute heart of everything that GWI is taking to market.

Speaker 1: Yeah. I, I love that. I love the idea of trends and looking for contrasts and, and so on. And certainly being able to do that with such a rich level of data over, you know, a, a very long period of time. Um, I think it leads me into next question, or kind of the first question, as it were. Um, you were on a GWI panel at Cannes Lions, and the theme of it was really around does Gen really exist? So let's start there. What was your answer?

Speaker 1: Does it exist?

Speaker 2: And in fact, I'm just... I've just finished a panel here at Adweek New York on almost exactly the same topic.

Speaker 1: Fantastic.

Speaker 2: It's, uh, it's a yes and no answer. Yes, technically, there is a cohort of people globally who are aged between, this year, 17 and 29 years old, who we could call Gen Z. But is there any coherency or unity within that group in the way that is often assumed? I would say absolutely strongly no. You know, at the, the lower end, you've got people who are single, living with their parents, probably in full-time education. At the other end, you have your 28, 29-year-olds, many of whom now are married with families of their own, in their own property, well into their careers. What really connects that 17-year-old and the 29-year-old, other than a label that the marketing industry has bestowed upon them as Gen Z? Not, frankly, very much. So yes, technically, but really, no.

Speaker 1: Hm. Could you say that Gen Xers are very different than Gen Z?

Speaker 2: See, what I love about-

Speaker 1: I'm a Gen Xer.

Speaker 2: Yeah. Well, actually, when I joined GWI, I had, uh, power to define the generations.

Speaker 1: Oh.

Speaker 2: So I could either make myself a very young Gen X or a very, very old Millennial. I am an old Millennial under the

Speaker 1: Yeah

Speaker 2: ... GWI classification. Um, yes, there are differences, but they're, they're gonna be driven by life stage and adoption of technology and mindset and value more than, um, the fact that you're born in one decade and that generation is born in another. What we see very strongly in the data is that your values, your beliefs, your mindset, your interests define you far more than demographic factors like age-

Speaker 1: Sure

Speaker 2: ... and gender, which are usually the, the building blocks of how much of so market-

Speaker 1: Right

Speaker 2: ... how so much of marketing, um, works. And so, you know, a good example, you could become a parent now at 20, you could become a parent now at 40. The act of becoming a parent unifies you with someone else who's been through the same thing far more than a, a simple marker like age. So I would always segment by something else first, interest, life stage, mindset, and then overlay age or generation to help you understand where to target, how those folks are consuming media, where they're hanging out.

Speaker 1: Yeah. Yeah. Um, I'm just thinking that, you know, that it's culture, right? So myself as a Gen X, there was acid house music, there was... Before that, other people who fall into Gen X would be punk, um, rock music, various other, you know, sort of musical influences. Uh, the art was different. Various other things were different at the time. There was no internet. Um, you talked about being, you know, the, uh, an older millennial. Um, perhaps you were born at the advent of, of the web

Speaker 1: There or thereabouts.

Speaker 2: Yeah.

Speaker 1: Yeah. Um, so it's always been there for you. I went to school, we didn't have computers. We didn't have phones. Various, you know, various other things that, that shape the way that, uh, you form your ideas, your opinions, and access information as well, more importantly.

Speaker 2: True. I mean, in my defense, when I went to school, there was one computer that was wheeled around

Speaker 1: Ha

Speaker 2: ... classroom to classroom, so uh, I'm not a complete digital native.

Speaker 1: Yeah.

Speaker 2: But yeah, I mean, uh, it is true that, uh, millennials, Gen Z, and now Gen Alpha who are coming through are digital natives, and that does shape certain behaviors and, and approaches to things in the way that earlier generations are playing catch-up on.

Speaker 1: Yeah. And the types of customers you work with, I assume fairly large global brands, agencies, you know, et cetera, what is it they're trying to understand about, um, younger individuals or, you know, perhaps old people like myself?

Speaker 2: Yeah, it's, it- it's interesting. The, the spotlight on Gen Z is phenomenal. I mean, if you look at the way people are searching for that generation, it outranks any level of interest that any previous generation has ever, um, achieved. Even millennials, who used to be the golden child, at the fever of sort of millennial obsession, we were never looking into them as much as brands are-

Speaker 1: Mm

Speaker 2: ... with Gen Z right now. Part of it is just the natural whichever generation is the youngest are perceived to be the trendsetters, forming culture, and broadly speaking, what they adopt first would then filter up to the older generations. But that would've been true of every generation that's gone. What seems to be unique about Gen Z, uh, is their, uh, the sort of digital native environment in which they've grown up-

Speaker 1: Mm

Speaker 2: ... and also that they were the youngest generation when we were going through once-in-a-lifetime events like COVID and, and everything that, that came from that. So, the desire to understand how they think, feel, and behave is stronger than it ever has been for any other generation as a result.

Speaker 1: Do you think that's because... I'm just thinking about media and the proliferation, the fragmentation of media over time as well, and it's never been so dispersed as it is now. Um, and also, I guess, sort of fragmentation of opinion and thinking and, and so on. Is that hard to measure and understand or am, am I going down the right track here?

Speaker 2: Think so. I mean, yes, fragmentation is, is there. If you look at the number of platforms, devices that, uh, especially the younger generations are spending their time across this, it's eye-watering. Um, and actually the, the total time that people spend online every day continues to go up. Yes, a lot of it's happening concurrently. You know, you're scrolling social media while you're listening to something or while you're watching something. So it's, it's, uh, concurrent rather than sequential, but still the amount of time that we spend per day online with our attention being given to things continues to go up and up and up. The problem is it's across a wider range of, of, of places and spaces than ever before. You know, if we were having this conversation probably five years ago, we'd just be talking about social. But now you'd be distinguishing social from short-form video, from long-form video, from various other things. Um, so the difficulty of reaching your audience across the plethora of, of places that they're hanging out is more, um, pronounced than it's ever been before.

Speaker 1: Mm. Um, I'm thinking when we were speaking earlier, you've obviously explained you've been at the business for the last nine years, and I think you said on day one there was around about 13 people in the business. You've obviously soared and, and grown and grown from there. Um, what was the, what were the outputs that you were providing back then, nine years ago? Was it a written report? Was it a survey-based rep- Was it, was it something else?

Speaker 2: Yeah. I, I, I mean, it's actually 12 years ago, so we're going even further back in time. When I first joined GWI, we didn't even have a proprietary platform. We were using a third-party vendor, um, to deliver our data in what now seems like a very static, clunky way.

Speaker 1: Mm.

Speaker 2: Um, you could interrogate it yourself, but it was, it was hard work.

Speaker 1: Yeah.

Speaker 2: Um, and reports, good old PowerPoint reports would've been a fairly substantial, um, offering. Go all the way forward to 2026, and there's still a use case for an expert research platform, and we continue to in- invest in all of the features that our power users want when they wanna go really deep in the, the data. But far more people are accessing via agentic interfaces, so via our own agent in the platform or via the connections we've got with ChatGPT, Claude, Gemini, Copilot. And the m- the major difference is that you used to go to the data, you had to analyze it, and you had to extract the insight yourself. Now the agent does all of that heavy lifting for you. It crunches far more numbers than you could ever have done.

Speaker 1: Mm.

Speaker 2: It analyzes all of the insights and surfaces what it thinks the, the strongest ones are. You still have to then apply your human expertise and context-

Speaker 1: Yeah

Speaker 2: ... to it-

Speaker 1: Yeah

Speaker 2: ... to understand which ones to take forward, but that's the biggest difference-

Speaker 1: Yeah

Speaker 2: ... I would say.

Speaker 1: So does that mean that you can test multiple hypothesis and do that at scale?

Speaker 2: Absolutely. I'll give you a great example. I asked the agent a couple of weeks ago to show me every place where consumers were contradicting themselves in their answers. So they say one thing over here, but then they express another thing over there that at face value is, is contradicting that.

Speaker 1: Mm.

Speaker 2: And I asked it not just to find all of them, but to then rank them in terms of what it perceived to be their commercial importance, and then suggest some follow-ups from it. Now, if I'd had done that as a human analyst, it would've taken me days, if not weeks. I wouldn't be confident at the end that I'd found everything that was a contradiction. The agent could do it in minutes.

Speaker 1: Mm.

Speaker 2: Yeah, uh, sure, it went over the 10-minute threshold, which is unusual these days, but even so, the, the time saving there, and the fact that it already ranked them for me, and had, in fact, given them pithy commercial names that I could have, uh, used to, to share them around the business-

Speaker 1: Yeah

Speaker 2: ... it's, it's just mind-blowing.

Speaker 1: That's interesting. So tell me more about the, the rankings. So I'm thinking, okay, you have knowledge, and from knowledge you get insights. Um, and those insights can be transposed into, you know, X, Y, or Z outputs, right? Um, how is the AI assisting with that? And, and tell me about ranking, what that means within those sort of scenarios.

Speaker 2: A, a lot of it comes down to the prompting, and I think as an industry we're still on a journey when it comes to what makes a good prompt versus, uh, what doesn't. And built into our product is some tips and tricks to help people-

Speaker 1: Mm

Speaker 2: ... uh, maybe think about how to, to rephrase it. So the output of, of any AI tool, I would say, is only as good as the data that goes into it and the context you give it.

Speaker 1: Mm.

Speaker 2: Uh, the agent has to know what you're, um, trying to achieve in any sort of, um, ranking. So the fact that GWI is serving the marketing and advertising industry is important context for a query like that.

Speaker 1: Mm.

Speaker 2: You have to say the end users who may need to act on this are, you know, these types of personas in these types of, of, of companies, because otherwise, with such a vast array of information, it could, you know, come up with so many different things-

Speaker 1: Mm

Speaker 2: ... but not relevant to the end user and their use case.

Speaker 1: Yeah. Does that mean as well that you're helping businesses, let's say it's an agency and they've got multiple clients, or perhaps they've got a long-term history, that might be a better way of describing it, with a individual client. Um, they can have huge amounts of data, some of it performance data, trends data, various other things within their enterprise. Plus, you obviously have your data sets as well. Are you helping them sort of liberate, I think is probably a good word, and, and free the data within their organization?

Speaker 2: Yeah. I, I think a good way of, um, thinking about it is the three intelligence types that can come together. You know, everyone now has artificial intelligence tools, whether it's a proprietary one you've developed yourself, whether it's one of the industry-leading ones. By and large, they can do the same type of thing. Then you have your proprietary intelligence, the things that you as a business or organization know about your customers and your product. It's all your internal data. GWI likes to think of itself as the third pillar, the human intelligence pillar that can come along and give the context to both of tho- to that central pillar using all of the capabilities of, um, artificial intelligence. The wealth of knowledge we have about, uh, how humans think and behave helps so many organizations out there give the context they need to their audience or their users.

Speaker 1: Mm. Do you also... Yeah, I, I like the idea of sort of predictive as well. Are, are you also involved in predictive outcomes in terms of how this audience will respond to perhaps, you know, X form of messaging or, or something else?

Speaker 2: We are, absolutely. So, um, there's the traditional way of doing that, which is fielding some, uh, new research among that target audience, or there's simulation. Um, we've been working behind the scenes to bring synthetic or simulated capabilities to market in a way that we can stand behind. I think out in the market at the moment, there's, there's a lot of vendors who are saying that they can do it-

Speaker 1: Mm

Speaker 2: ... but it gets a bit smoke and mirrors, and personally I think some of the confidence metrics out there are slight red herrings.

Speaker 1: Mm.

Speaker 2: We've been working really hard on simulations where you can see how the entire audience has, has answered. So there's not just one headline metric, you can see how every individual would've answered. And because we have 15 years of, of data, we're sitting on a goldmine of the best possible type of data, because our testing would show that the best way to predict how someone would answer is looking at how they have previously answered to related, um, questions. So yes, we're entering the predictive space, and from next year our platform will start doing far more in terms of, uh, intelligence segmentation predictions-

Speaker 1: Mm

Speaker 2: ... or spotting the trends that are emerging before you've even thought to ask for them.

Speaker 1: Yeah. Yeah. I, I'm, I, I'm j- I'm just thinking a little bit deeper around the idea of synthetic panels, synthetic data, et cetera. Um, do you feel that there's still a, a bit of fear and hesitancy, uh, in terms of using that data? I ask this for a reason. I can't name the client on our side, but we also own a research business, and within the master services agreement, they've written into it that we should not ever be using synthetic data in anything that we do with them at this point in time. Um, I look at it slightly differently, and actually I f- I feel that synthetic data is potentially more authentic than, um, perhaps running a survey or multiple surveys or through different panel providers. So what's your take on that? And, and actually, how should I have perhaps... Perhaps I go back to them and, uh, overcome that objection.

Speaker 2: Yeah. I, I think if we were having this chat a year ago, I would've agreed that there was a lot of fear and uncertainty.

Speaker 1: Mm.

Speaker 2: I'd say right now that's given way to outright skepticism, um, about, uh, whether the data is trustworthy or, or usable.

Speaker 1: Mm.

Speaker 2: I think it's like anything in the AI space at the moment. If you put garbage in, you'll get garbage out, and there's too many people, in my view, who aren't really understanding the right source data, the methodologies that you can employ, and just saying, "We can predict X, Y, Z" Well, actually, I think the data is not there to do it. I think synthetic respondents can be excellent, but they can also be very limiting and biased, and it's all in the context and what you're trying to achieve.

Speaker 1: Mm.

Speaker 2: And generally speaking, I would say that if you need a quick steer, if you're early in a process, if you need to narrow down options to a shortlist, those are great places to consider, um, synthetic responses as long as you're confident you've got the source data that can, um, inform it. If you're making huge budget-influencing decisions or things that will define your strategy, you probably want to be a lot more cautious right now.

Speaker 1: Mm.

Speaker 2: But wherever you are, really scrutinize the vendor, the capability, and the claims they're making because people in this space need transparency, and they're, they're just not-

Speaker 1: And it's, it's-

Speaker 2: ... getting it at the moment

Speaker 1: ... lacking at the moment. So what you're saying is there's also a little bit of snake oil going around at this point in time?

Speaker 2: I think so. I mean, there has been an explosion of capabilities in the last year.

Speaker 1: Yeah.

Speaker 2: Almost like a land grab for who's gonna, um, establish themselves.

Speaker 1: Mm.

Speaker 2: And so many of the clients that we speak to each week are a bit exasperated-

Speaker 1: Mm

Speaker 2: ... of not being shown what's inside the black box, um, and yet being told these really high claims that it's 95% confidence this is right. And you know, there's a great example that I looked at last week where, um... not on GWI, I should say. Um, we'd given it some source data that was, uh, you know, tell you about someone's digital life, and then I asked the, the, the agent, "What's, what are their favorite flowers?"

Speaker 1: Yeah.

Speaker 2: And I gave a list of five flowers, and it came back very confidently saying, "This type of flower is this audience's favorite flower." There's literally nothing in the source data that could have informed that.

Speaker 1: Mm.

Speaker 2: But the agent that was doing it was very confident, um, in its answer and said, "You know, I've run it multiple times, and every time, it was the same flower at the top of the list." It's like, good for you, but I don't have any belief that it's founded in anything that I can, I can believe, and that's-

Speaker 1: Yeah

Speaker 2: ... that's the problem. It, it... the overconfidence of AI-

Speaker 1: Yeah, I agree

Speaker 2: ... comes out very clearly in synthetic-

Speaker 1: Yeah

Speaker 2: ... uh, responses.

Speaker 1: Or it's almost trying to re- please you sometimes and give you the answer that it believes that you want to hear.

Speaker 2: Oh, completely. I mean, whichever tool you're using, just take that moment to, to stop and say, "This isn't good enough" or "You should have done that." And, and almost certainly it would say, almost straight away it would say, "Yeah, good point," or, "Yeah, sorry, I..."

Speaker 2: So, yeah. But those who don't do that-

Speaker 1: Yeah

Speaker 2: ... are just getting the, the, the initial confidence, and most models were trained to affirm the views-

Speaker 1: Yeah

Speaker 2: ... of the user that was inputting into it.

Speaker 1: I had to tell my GPT off yesterday. "GPT, you're meant to be my life pilot. You've done this, and this wasn't quite right." And it said, "Well, John, I gave you this answer. This was the qu- these were the questions you asked. Yes, I did know about this as well, but I didn't provide that information to you. I'm really sorry. Next time I'll expand on things a little bit more."

Speaker 2: Yeah. No, I've been exactly... I regularly have arguments with AI tools that

Speaker 2: ... are not quite delivering what I want them to.

Speaker 1: Yeah. Um, what are the biggest use cases that, um, customers are coming to you with?

Speaker 2: Ultimately, everyone who is using GWI needs to understand something about their audience. Maybe that's their existing audience, how it's changing, what impact various pieces of technology or, or world events are having on them. Maybe it's target audiences that they're going after in, in the future. Maybe it's competitor audiences. It... really, it's understanding not just what an audience is doing, it's why they're doing it.

Speaker 1: Mm.

Speaker 2: And that's where I think survey data really comes into its own because you get people's perspective, their view on why they're doing it, and all GWI data is volunteered by people, so they know what they're disclosing. They're happy to, to give that, um, data, and I just find there's such richness in people's perceptions and their motivations that give the context for all of the behaviors that you see.

Speaker 1: Mm. Mm. And do you go beyond the panel, as it were? So in terms of how you source the data, how you get the, the respondents, the, the, the, uh, the response to particular questions?

Speaker 2: So we work with, uh, about 30 different panels, um, so make sure that we have the widest, most representative, um, sort of pool of respondents to draw from.

Speaker 1: Yeah.

Speaker 2: Um, we do, uh, occasionally go to other sources as well, but from panels is the main place that we're, uh, bringing all of our intelligence from.

Speaker 1: And what's the split between B2B and, and B2C?

Speaker 2: In terms of our customers use?

Speaker 1: In, in terms of your customer base, yeah.

Speaker 2: Um, it's a great question, and it, it sort of varies week on week, depending on, on, on what projects-

Speaker 1: Yeah

Speaker 2: ... uh, are coming in. We know... We work with, um, all of the leading agencies, and so they would have a very diverse, um, client base, so that influences it. We do work with some B2B customers. We also work with many B2C customers, so there's, there's no one answer I can give you there. It's, it's, it's both-

Speaker 1: Yeah

Speaker 2: ... coexisting.

Speaker 1: Yeah. I'm just wondering if there's... if you're seeing growth amongst B2B, um, certainly for synthetic audiences and, and various other things, uh, that perhaps they haven't come to you around in the past.

Speaker 2: I think when it comes to synthetic, it's, there's growth in any hard-to-reach audience, so B2B is one of them.

Speaker 1: Yeah.

Speaker 2: But you would include, um, you know, healthcare professionals-

Speaker 1: Mm

Speaker 2: ... or high-net-worth individuals or, in certain countries, particular demographics. Um, uh, you know, synthetic is a, a great way to give you reach that you might not be able to do if you're reliant on speaking-

Speaker 1: Yeah

Speaker 2: ... to, to people themselves.

Speaker 1: Yeah. That's exactly why I asked the question, so obviously, you know, the dream. We... Can we reach 1,000 C-suite with these questions? Why are they interested in your questions? Well, how are they going to respond? How, how are they incentivized? Are they not? They say, are they not really busy, got better things to do? So I, I was just thinking, yeah, but actually, that's probably of quite a good use case in terms of at least getting a steer and, and getting some initial insights from synthetic.

Speaker 2: Yeah, absolutely. But again, it comes down to you need some initial input. So if you want 500 CEOs, you're gonna have to speak to a decent portion of them to inform that, and you will have to pay them far more highly than if you were speaking to your average consumer who... on the street, as it were.

Speaker 1: Yeah, of course. Um, what are some of the more interesting questions that people have come to you with recently?

Speaker 2: Well, well, I suppose in the agentic, the brave agentic world we find ourselves in, my favorite use case is, uh, tell me something unexpected or surprising about XYZ audience. I do it all the time now, 'cause, A, it's a great way to show the value of the, the product.

Speaker 1: Yep.

Speaker 2: But it's also a great way to challenge thinking among whichever user you're, you're working with, 'cause all of us have our preconceived notions of what our audience is like. Dare I say it, sometimes we're susceptible to stereotypes. In a few seconds, our agent can look across the entire breadth of all the data that we have and tell you the things you weren't expecting or which seem counterintuitive. And I think the, the golden rule of, I was gonna say everyone in the insight world, I can't speak for everyone-

Speaker 1: Mm

Speaker 2: ... but probably most insight professionals is you never want insights that are bland or expected. It's the surprising, challenging, unexpected one that is often the biggest light bulb moment that can lead to a, a new campaign or new product development that actually moves the needle for you.

Speaker 1: Mm-hmm. Yeah. How has... I think you've answered it to a certain extent, but yeah, how has the proliferation of AI in your business transformed fun- fundamentally the purpose of your business?

Speaker 2: It... Yeah, and it, it continues to.

Speaker 1: Yeah.

Speaker 2: Uh, you know, we're not at the end of that journey yet. But what AI has done for us really has allowed us to democratize insight. So in the old approach, as it were, you would work with a relatively small number of expert users, all of whom would need a research background, and need some training and onboarding, and they'd need to keep themselves coming back to the platform so they didn't forget how to do stuff.

Speaker 1: Mm.

Speaker 2: They'd need to go and find the insight themselves. Now, we have organizations where hundreds, thousands, or tens of thousands of users have access. They're not from a research background. They don't have knowledge of what GWI has collected or how the research works. They're just able to ask the question and receive the answer. And AI has been revolutionary in that in terms of taking us from data to insights and from the specialist user to the mainstream, uh, any persona can get value immediately.

Speaker 1: Mm. Yeah. Yeah, I totally agree. I think the, the democratization aspect's incredible. Really is. Um, and actually, this shows a good example. So we've recorded about 180 episodes, 150 or so have aired now. Um, I look at the, I guess, the body of work, the transcripts that we've created, um, and the corpus material that sits there now, and ask myself the question, are we at... Could we be at the stage where we could answer some fairly profound questions that a board of directors might have? Um, one of those could be, "Well, over the last year, we've lost half the traffic coming to our website. Why is this? What, what's happening? Can, uh, someone explain it? Help." And those answers are there-

Speaker 2: Yeah

Speaker 1: ... from the aggregate of all of the conversations that, that we've had, uh, with various leaders from, from a whole array of, of businesses. Um, looking at an industry like retail for, for a, uh, instance, I created a report recently about the idea of the, the last store standing. Um, that was informed by a conversation I originally had with Dimi Alpers, who's the CEO at Depop, and that was at, at Cannes Lions. But it wasn't just him. We were talking about, you know, what, what becomes removed from a business as, as AI, you know, obviously increases. Um, and that could be part of its, its value. It could be, you know, uh, how it might end up transforming, you know, a whole myriad of things. Um, and there were a number of conversations around that, and then there was conversations with retail businesses and, and so on, which has enabled, you know, the production for a, a layperson, I guess. I'm not a, an analyst by trade, to be able to produce quite a high-quality piece of content, but more importantly, with the answers that people are looking for. So yeah, I, I, I totally believe and I, I really love the idea of, uh, democratization of, um, research, um, of insights are things probably a, a, a better word, and making insights much more accessible to people within the business.

Speaker 2: I, I think it's not just that it's democratized, although that is super important, it's democratization with verification.

Speaker 1: Yeah.

Speaker 2: So any user in our system can immediately verify the insight and trace it back to real humans who said this thing-

Speaker 1: Yeah

Speaker 2: ... in this country. And I think that need for transparency or, uh, reporting is gonna go up and up and up-

Speaker 1: Yeah

Speaker 2: ... because, uh, you know, we've surveyed professionals, 99.9% of us are using AI, and that's not going anyway-

Speaker 1: Mm

Speaker 2: ... anywhere. The tools we use might change, but, you know, the genie's out the bottle.

Speaker 1: Yeah, yeah.

Speaker 2: AI is part of our work streams forever now really. But people needing certainty over whether that figure or insight is correct and verifiable-

Speaker 1: Mm

Speaker 2: ... is extremely strong, and I think-

Speaker 1: Yeah

Speaker 2: ... will get more and more attention.

Speaker 1: Yeah. That's also, you know, I'm gonna use the word democratize again, but democratizing access. So I'm gonna make the assumption that your user, as it were, in inverted commas, looks very different than the user of the past. Perhaps you had a head of content or a research director or someone else who was y- your main client. Uh, now I'm assuming that multiple individuals would... In, in a team are accessing your data and insights for a whole myriad of, of reasons and purposes.

Speaker 2: Yeah. I mean, the, the research experts or professionals are still there. Um, and they might over-index towards using our, um, platform, where you can go deeper and do a lot of the analysis-

Speaker 1: Sure

Speaker 2: ... yourself. But they're using the agents as alongside that. But they're now joined by all of those personas that previously would've been solely reliant on the research experts that can now get in there and, and self-serve what they need.

Speaker 1: Mm.

Speaker 2: So, you know, your head of comms, your head of marketing, your sales directors. Really every job title you could imagine throughout the marketing and sales hierarchies we see showing up and, and accessing the insights, um, on a kind of self-serve basis.

Speaker 1: Yeah. And I guess they've got different, uh, ways that they wish... You know, you, you talk about the idea of just asking a natural language question, for instance, versus, um, using tools to better slice and dice data and, and be more sophisticated, perhaps if you're a research director. Um, so I'm guessing that, um, based on sort of different individuals, their particular use case, the way... Actually, the way that they probably want to digest and receive information that, yeah, that I would imagine that that's quite a broad spectrum and, um, yeah, maybe telling o- of different people within different areas or categories of business and what they're looking to get out.

Speaker 2: Oh, yeah. Totally. And it, it's great learning for us as well because it shows us how we need to carry on evolving our product and the data that w- we're collecting. 'Cause you're quite right, different personas have very different, um, needs but also different types of questions-

Speaker 1: Yeah

Speaker 2: ... that they're, um, putting in. There are different types of support we need to be, um, providing as a result. So it's, it's an amazing goldmine for your own product development.

Speaker 1: Yeah. But, um, are there particular brands at this point in time that you admire in terms of how they go to market, how they show up, how they represent themselves, and how they market to their customer base?

Speaker 2: Well, GWI's in the fortunate position that we work with almost all of the the names who are on stage.

Speaker 1: Yeah.

Speaker 2: So I, I, I need to be careful about calling anyone out. But, um-

Speaker 1: Or just a brand that you admire. You know, it could be that they're purpose-driven, something else, um, that you buy into their marketing.

Speaker 2: Ooh, that's a great question. Um, I actually think... I'm not a customer myself, but I think they've had an amazing, uh, journey of adoption. The sneaker brand On, uh, does a really good job. And so many-

Speaker 1: Mm

Speaker 2: ... of my... It's always telling when your own friendship group starts one by one-

Speaker 1: Yeah, yeah

Speaker 2: ... um, becoming, uh, customers, right? And I think they've, they've done a really good job at disrupting their particular, um, uh, the space that they're in against the legacy players and creating that sort of buzz and exclusivity even though it's a mainstream product.

Speaker 1: Yeah. Well, they literally came from nowhere overnight-

Speaker 2: Yeah

Speaker 1: ... as it were, um, and exploded.

Speaker 2: Absolutely.

Speaker 1: And yeah, I'm not a customer yet. I'm not sure I'm gonna become one now, but, uh, a couple of years back, I, I was seriously in consideration. Um, a go-to resource. You know, how do you sharpen yourself up? Um, is there a particular information resource or perhaps just a place that you like to go to, um, gain ideas and inspiration?

Speaker 2: Yeah. I mean, staying on the sort of conference circuit, for want of a better term, is, is crucial. Over the last year, I've been to South by Southwest, Adweek London, here now in Adweek New York, uh, CES possible. Obviously, Cannes Lions.

Speaker 1: Yeah.

Speaker 2: Um, you really can't beat, uh-

Speaker 1: No

Speaker 2: ... some time in the French sun while you're nonetheless improving your knowledge of what's going on.

Speaker 1: Yep.

Speaker 2: I think it's really helpful to see what people are showing up talking about.

Speaker 1: Mm.

Speaker 2: Uh, what are the majority of topics focused on? What are people's questions and concerns? Y- you know, we've talked about synthetic already, but you can see that showing up in, in one of the big discussion topics this year. People looking for guidance about how to go about it, but also a bit of certainty from vendors. Um, and it, it helps us at GWI as well because often clients are looking to us for thought leadership of, how do we go about adopting this or improving our systems? So I will always make time, uh, to come to Adweek or any of the other, the big, uh, names each year to-

Speaker 1: Yeah

Speaker 2: ... to make sure that we're understanding what everyone's talking about.

Speaker 1: And then lastly, advice that you may pro- provide to a younger individual, perhaps at the start of their career, uh, or looking at the world of, you know, marketing, media, advertising, um, research perhaps. What piece of advice would you give them at an early stage?

Speaker 2: I'd probably give two pieces of advice. One that's evergreen and one that's a bit more of the moment. The evergreen one is always look at what the data's telling you before you form your opinion. Too often you come to the data with your preconceived idea. You don't let the data tell you its actual story. You look for the story you want it to tell-

Speaker 1: Mm

Speaker 2: ... and you, you carry on from, from there. So always make sure you're being neutral. But then the other one, which is more of the here and now, is there's so many people questioning, what does AI mean for this, um, this career, this profession? Right now we're hearing a lot about human in the loop, keeping the human context. So really think about how you can develop your skills and your experience that you are essential alongside the AI tools that are being adopted. Now, I firmly believe that there is no substitute for human experience, human context. You need to be the one that's refining, curating, telling that story. But therefore, you need to be quite bold on your... Or confident in your abilities and the way that you present that-

Speaker 1: Mm

Speaker 2: ... because, uh, you know, AI is changing the game. It is, uh, it's shifting the parameters within which we do our jobs. So think about how you remain relevant, not just relevant, but necessary and important within that.

Speaker 1: Yeah. I couldn't agree more. Fantastic advice. Thank you so much, Jason, for joining us on the show today.

Speaker 2: Thank you for having me.

Speaker 1: You're very welcome.