Jamie Credland  0:00  
For the first time in many years, this is a new technology that's coming into the market where younger people trust it less than older people, and I think from a marketing perspective, and we've all been in the industry long enough that you know there was there was the web, and then there was mobile and apps and social. There's a new thing that comes along, and the whole industry races to embrace it as quickly as possible. Normally, because younger consumers and future consumers are crying out for brands that are active on mobile or vertical video or whatever it is, that doesn't seem to be the case right now. Where, in fact, younger consumers are pretty sceptical about brands who are visibly using AI?

Ben Walker  0:55  
Welcome to the CIM Marketing Podcast. This is the first in a two-part series examining CIM's latest research on marketing in the age of AI with none other than YouGov, and I'm delighted to say that from YouGov itself we have none other than Miss Jemma Conner, and Jemma is an associate director in the political and social research team at YouGov, and she's mainly responsible, actually, at the pollster for pan-European polling, but she also leads on Welsh polling. And I have to say, in the recent Welsh elections, the Senate election, she was the most accurate of all pollsters. Gemma, great to have you on the show. How are you?

Jemma Conner  1:37  
Hi, I'm very well, thank you. Thank you for having me.

Ben Walker  1:39  
It's great to have you on the show. Great to have you on the show, and the great hot guests do not end there because we also have with us today mr. Jamie Credland. Now, Jamie, many of you will know he is, of course, chief executive of the World Media Group. Before that, he spent 15 years as a marketer at the Economist, and before that was agency side at UM. Jamie, great to have you.

Jamie Credland  2:03  
Great to be here. Thanks, Ben, for having me.

Ben Walker  2:05  
Fantastic to have you on the show. And last but not least, many of you will know mr. James Delves. Now, James is head of PR and external affairs at the Chartered Institute of Marketing. He's of course a chartered marketer himself and a CIA fellow. He brings more than 20 years of experience leading agency and in-house marketing teams. But most of all, he's back on the show after what is a prolonged absence, mr. Delves. We're trying to count up the years, but it's great to have you back. I'm going to ask you, James. You know, AI has transformed the marketing landscape. We know we're all using it. Why did you do this research now? Why launch this report at this moment?

James Delves  2:48  
Just felt the perfect moment, Ben, because marketers have kind of moved from experimentation to widespread deployment. We're not as clearman prompts more AI is now hardwired into marketing teams, but that acceleration has really created a real auditory paradox. Bulls everywhere looking at Gen AI through quite a narrow lens at the moment: cost reduction, margin improvement, and in some cases, real headcount precision. So it just seemed the perfect time to look at some of those really big issues, drill down of the data with UGov, and to come and really provide marketers with a kind of where we are now or where we want to go in the future.

Ben Walker  3:28  
Jemma, cost reduction and headcount reduction are not in phrases that generally meet well with the ears of marketers. Can you tell you what you found in your research on those issues, particularly?

Jemma Conner  3:41  
Yeah. So overall, AI is being used for a variety of purposes by marketing decision makers. If we look at the sample of that group, around a quarter are using AI to improve operational efficiency on the non-creative side, with similar proportions using AI to take a closer look at their competitors and track market trends. Around a fifth are using it for price optimization. However, when we looked at how AI is being used and actually the extent to which it's being used to achieve their ROI, it does seem to depend on the size of the organisation. Large companies with more than 250 employees are far more likely to be using AI to achieve their return of investment than smaller companies. For example, 53% of marketing decision makers working for larger companies say they're using AI for price optimization, compared to just 15% of those who are working for a smaller or medium size enterprise.

Ben Walker  4:37  
That's interesting. What else in there did you find that you think is a big headline finding in terms of the usage of AI in the sector?

Jemma Conner  4:45  
It's interesting how many businesses do not expect to achieve any return of investment from using AI before 2030 It's 40% of marketing decision makers are saying that that's not going to happen.

Ben Walker  4:57  
40% are going to use this before 2030 They're not going to expect a return since 2030

Jemma Conner  5:03  
Absolutely, and that increases massively when we look at smaller companies. So, 52% of micro companies or small companies, those with less than 50 employees, say they're not expecting that return on their investment. And 47% of small and medium enterprises, whereas large companies, those with 250 employees or more, and just 1% are saying they're not expecting to see that return. So there's a real difference depending on the size of the firm.

Ben Walker  5:31  
Real difference, James. Big findings there. The size of the firm is making massive difference in terms of the expectation of ROI, and indeed the size of the firm also seems to make a big difference in terms of what they're using it for. These dread phrases, cost reduction, these dread phases of headcount reduction. You've done a series of roundtables on this very matter. We met organisations, you've met senior marketeers. What were their reactions to these findings?

James Delves  6:01  
Yeah, it's quite interesting. So we obviously did the research with I'm German's team, and we really wanted to stress test it to basically take it out and find out what people in ordering, but what way people in the front are what they actually felt of the findings. So we held two roundtables with IBM in the renovation centre in Central London. They were attended by between about 30 and 40 market professionals, everything from seniors TMOs at global banks all the way down to IBM's AI team, sim course directors to give an academic view. We had young interns in there, and then we had senior marketers kind of operating on the front lines. And you could just see people's how the energy changed in the room, when we put the figures up on the screen and we discussed them as the team, we laid all the figures out. Would everyone really dig into them? And there's a real consensus that AI is splitting marketing into sort of two distinct operations there in the front office. So the back office, everyone was really positive. Great, it's going to take away all my effort around my mixed modelling, my automation of internal reports, all my data sets-they were really positive. You could physically see people's faces light up when they could see that the cogs whirring, and I'm going to clear all this rubbish off my plate. And then when we moved it to the front office and started talking about things like customer-facing copy, creative execution, brand storytelling-you could just see the room got a bit quieter, and then people were distinctly more cautious. And we try and not conversation about things like guidance and legal and how we would do it. So I think a lot of people, when they looked it through, could see the massive benefits, but also they could see some of the kind of the warnings, such as like 60-3% of adults lose trust in a brand when they see core AI content. So a lot of them select a really big wake-up call. It's not the big silver bullet where they can just load a whole bunch of content into it, drive productivity and output. Actually, they need to make sure it's really of high quality. Otherwise, people now are getting susceptible to that, and are particularly looking at some of those content and going, you know what? I'm going to step away from that brand because I don't think they value me, or the content they're putting out isn't good. So there is a bit of a trap there we need to watch out for. So they were the kind of things that came out. It was really interesting how it really quickly split the two sections. How it was going to be used completely chains people's physical response to

Ben Walker  8:21  
the data. It's absolutely fascinating, Jamie. You've been nodding furiously during James's testimony. The pace of change has been incredibly rapid, and you've probably spent most of your time at Cannes this year discussing that pace of change. Are we adapting to this stuff properly?

Jamie Credland  8:37  
I think. I mean, it's the perfect time to be talking about this report. I think, as you say, I was at Cannes Lions last week, and WMG World Media Group. We spend a lot of time dealing with advertisers who tends to be at the larger corporate advertisers, the larger B 2b kind of brands. And there is there across the whole festival of and conference in Cannes. There was a real noticeable shift this time, 12 months ago, it was unbridled optimism about AI that every single every single process and skill within the marketer's toolbox was going to be replaced. And yeah, in in fairness, in the last 12 months, marketers really have embraced this. I didn't. You don't meet any marketers who aren't using AI in some way within their workflow. Many people have been encouraging their teams to adopt it, and you know people are very excited about this technology. But I think perhaps there's also a little bit more realism cooking in about which tasks AI is really fantastic at, as James says, some of those back office tasks, and which tasks AI might have some value to, but it's not a replacement for human oversight. The more front office pieces, and I think this additional thing, which is is highlighted in the report, but it really is coming out many many places in many different categories and industries. Again, this piece around trust and truth is that what you do in the back office. I think often consumers are not potentially that interested in how you do your media mix modelling. That that's fine, but as soon as you're putting something in front of a consumer that is obviously AI generated or touched by AI, I think either you have to be transparent with the consumer, or there's a real risk to trust if it's not clearly labelled, and the other piece around that, which I think is also in this report, but for the first time in many years, this is a new technology that's coming into the market where younger people trust it less than older people, and I think from a marketing perspective, and we've all been in the industry long enough that you know there was there was the web, and then there was mobile and apps and social. There's a new thing that comes along, and the whole industry races to embrace it as quickly as possible. Normally, because younger consumers and future consumers are crying out for brands that are active on mobile or vertical video or whatever it is, that doesn't seem to be the case right now. Where, in fact, younger consumers are pretty sceptical about brands who are visibly using AI. So, there's a lot of enthusiasm, but there's also a certain amount of healthy scepticism. I think, which wasn't there previously.

Ben Walker  11:16  
It's interesting you raise that about the generational aspect of scepticism, because I was reading this report last month, I was looking to get a hold of an embargoed copy from CIM prepublication, embargoed copy, locked down version, and that was the thing that jumped out to me that the younger generation, if you like, the Gen Z Zoomers, were the most sceptical of of all of the generations of this stuff, and we're almost sort of-you got the impression that they were sort of almost looking out for AI Gemma, and if they could spot it, would then walk away from the brand.

Jemma Conner  11:54  
Yeah, this group is actually really interesting because in polling, we tend to see that the youngest age group, so in this case, the 18 to 20-four year olds are normally most likely to say they don't know about something, and actually in this polling, they're the least likely to say that, and they have very strongly held views about AI and how the use of AI in advertising and marketing really impacts their trust and their view of the brand. So across the whole nationally and politically representative sample of 2000 adults in Britain, 63% felt that they would lose trust in a company that was using AI generated content. That number increases to 79% amongst 18 to 24 year olds, and which I think is surprising. I think people kind of grow up in that digital age, and we kind of expect them to be the most savvy with AI and to be perhaps most on board with it. Actually, eight in 10 of this group say they'd lose trust in a firm, which is just a really interesting finding.

Ben Walker  12:56  
When I read this report yesterday evening, I sat there with a cup of tea, and I thought there's an element that there's an under there's an undertone here, a little bit of a bandwagon jump by this industry, probably by other industries, getting on this train before we've got the right drivers, getting on the train before we've got safety measures in place, almost like the technology, the train is getting ahead of the people it's supposed to be carrying. Is that fair analogy? Do you think?

Jamie Credland  13:27  
I think that's probably the case in some different ways. And like I say, I think previously the message from senior marketers to their team, because there was some concern about AI use within marketing, a lot of senior marketers were just saying that to you. I don't care how you use it. Just make sure you're using. Get on that. Get on ChatGPT every day. I want you to be using it for everything, and that's a completely reasonable response because you want to develop these skills within your team and get people comfortable with it. I think that's very sensible. I was talking to a couple of people at large professional services firms over the last couple of weeks with global operations who have been taking that kind of approach and have now suddenly realised they have teams in Europe and teams in the US building things that are completely duplicative. Right, they're completely wasting energy because everybody is doing all this stuff on their own, and there hasn't been a central organised process of okay, these are the projects we're going to use AI to deliver, and these are the teams that are going to work on those specific tasks. This comes out in the research report as well, but we're talking about a technology. This technology isn't going to replace the need for a marketing leader to structure a team, set clear objectives, set priorities, and at the moment, it's great fun. Everyone's gone off and built all kinds of widgets and bots, and who knows what what else. But especially if you're a larger organisation with 1000s of people, you need to at some point make sure everyone's on the same systems and something from the same handshake.

Ben Walker  14:56  
It's interesting, Gemma. That that that leads me to this other thing that I. Potted yesterday when I was reading the report that there seems to be a large discrepancy between those who are expecting a return on their investment versus how many of them have actually got a strategy in place. You know, you could almost call it a strategy gap, couldn't you, from by looking at the research?

Jemma Conner  15:16  
Yeah, absolutely. So as we were saying earlier, there's 40% who say they're not expecting to achieve any return on their investment from using AI in the certainly in the next few years. However, among the same group of marketing decision makers, almost half are saying that their companies not developed a formal strategy to address the new skills that are needed to use AI. So maybe that's why they're not going to see that return. They've not made that investment in the right areas of the business to understand actually how AI can lead to return in investment. It is a much bigger problem again for smaller businesses, and just 4% of decision makers at large companies say they don't have a strategy in place yet. Amongst SMEs, it's 50-5% and 60% of those working for micro or small companies, so they just don't have that policy. And yet,

Ben Walker  16:04  
James, it's another trap, isn't it?

James Delves  16:06  
Yes, basically, and it was something that came out loud and clear on when we talked to all these ex. I mean, you've probably heard that you know J.P. Morgan are spending more tokens than they spend on salaries. You know, Uber exhausted their entire AI coding budget by April. You know, Microsoft have basically revoked all their claw code licences because of sparring costs. It goes back to Jamie and Gemma's point: is that there isn't the the AI skill strategy in place in lots of firms, and it may they may well have an overarching one. But what does that mean for department level? What does that mean for marketing? And so there's that came through an awful lot. Is that you could say we are sleepwalking into an issue at the moment, and it's the gap between approving an AI budget and actually driving AI strategy. That's where the issue is having. As Gemma said, like 40-7% of decision makers and who were bought said they had no formal AI skill strategy in place. Everyone's chasing an immediate return, and sometimes a restructuring roles without reskilling people. And the bottom line, it hits twice. It hits first where it's commercial risk, and then there's the AI trust penalty, which Jamie talked about earlier. If we automate everything using AI, then the entry level roles are reduced. That that's basically where the staff seem to be getting cut from. But then it's where's the next generation marketers going to cut? How are we going to learn their trade to then move up the level, and how are they going to develop that that you know the judgement when the AI is telling them stuff that that's not actually correct? And if we cut the bottom of the pipeline, then the bottom line suffers. Just you know, or three years down the line. So I think we you could say we're sleepwalking into it. I would also say that you know this does sound very negative, but this technology wasn't around 10 years ago. So everyone's kind of scrambling, upskilling, employing it. And there is some really good stuff that brands are doing, yeah. It's just a watch point. We need to make sure we empower the staff, give them these new ways of working in a bit of a controlled, safe environment, but with the guidelines in place. Because otherwise, back to Gemma's point is, you know, without central strategy, we are going to base AI resource.

Ben Walker  18:17  
Got on the point of the strategy gap, which I think is a real worry. There's actually a quote in the report itself from Sam Winston of Salesforce. I read it now. It's quite an interesting quote. I thought he says the real strategy gap isn't about tools, but it's about AI fluency, the ability to embed agents or AI agents into a team's culture to reclaim the time to create and focus on high-impact human judgement and innovation. We've talked about the pipeline, but what does this AI-fluent marketing look like to you, James?

James Delves  18:53  
Well, I think it's probably the antidote to the problem I just discussed. If you can build a AI-fluent team, it's going to look less like a factory assembly unit where everyone's been bits, and more like a high-performing newsroom where AI is running around acting as like junior researcher, shifting data, spotting trends, handling programmatic buying, sending stuff back to you, sending drafts back to you, and then the humans don't step back; they step up into the editor, kind of roll at the very top, and then they could use machine intelligence to do back office stuff and give themselves time to create, and that really highlights the value I think of humans in the AI process. But you've got to shift that kind of that cultural internally as well, because you can create infinite amount of AI content, cheaply, but it's the human creativity that gives you the USP. Yeah. So I think that building those AI fluent teams is going to be difficult, but that could be the way forward where we really see the real benefit of these AI tools.

Ben Walker  19:59  
Some AI tools. Are being used more prevalently than others. There's some nuggets and researches on the Gemma about what the industry, what tools the industry is using.

Jemma Conner  20:08  
So mainly, people are using the backroom stuff, the things that we've been talking about, the operational efficiency, the the tracking what's going on in the market, and also the price optimization. But how it's being used does depend on the size of the company. Those larger companies are more likely to be using it for price optimization or yield management, and they're also more likely to be using it to to make their operations more efficient. And those smaller companies are much more likely to have not not really worked this out yet. It's the larger companies that are really leading the way in terms of using AI in their background workings.

Jamie Credland  20:46  
It's interesting, isn't it, Jamie? That price optimization for the bigger firms-these are the firms with 250 plus employees-is the most popular use. Not a fairly close second is the operational efficiency. Think that the marketing industry historically loves an easy to measure metric, and then tends to not worry too much about whether that's the right metric to be measuring in the first place. And and we've seen this happen in the world of online advertising with you know clicks and impressions become totally meaningless metrics. But the entire industry obsesses about. I think likewise, you know AI is a wonderful tool. I don't want to sound like I'm not in favour of using it across the business, but if the only metric is cutting your costs, I mean, the the absolute best way to cut your costs is don't do anything at all, and that and that's guaranteed to work. So, a you know, you you lose a lot of the creativity and the specialness of what makes marketing. The other thing, though, is I heard this new expression last night from someone at PwC of tokenomics, the economics of tokens. But I think, as James was getting at, I met some people last week who had reduced their team of web developers by I think 10 to 15 people because of supposed AI cost savings, and we're now spending more on tokens than they were on the staff. You know, it seems like it's a really quick win to just get rid of junior staff. And as James already said, there is a question that raises for the pipeline for future talent, which is really important. But also on a just purely financial basis, it's not obvious actually how much these AI tools are going to cost, and also, however much your tokens cost today, is probably the cheapest they're ever going to be, because as marketers build these relationships with 234, major LLMs, right? And there's not that many players in the market. We've seen it before. You know, think about Microsoft in the 90s. Once we were all on Microsoft Office, no one was leaving it. And there is a point if you build your entire business around systems that are delivered through Claude. When Anthropic, understandably, say we're going to increase the token cost by 10% your business as well as will negotiate in power. Whereas having people within your own team, you can always move around the tasks that they're doing. You can always scale them up and down. You have a lot more control as a marketer. So I think it's interesting that there are absolutely tasks, and I think you know pricing analysis, very heavy data pieces. This is exactly what AI is designed for. I think some of the junior tasks are perhaps not as much of a cost saving as people think, and then finally, I think we we shouldn't forget what AI is, which is a really really fantastic tool to predict the next word in a sentence. By its nature, it regresses to the mean. That that is literally what it's designed to do, and that's fantastic for for many many things, but if you are looking for you know creating a tone of voice for your brand that is provocative and different from everything else in the market, or you're looking for a creative idea-not I don't just mean in terms of a 32nd TV ad, but in terms of a creative idea for your strategy-there's a real need for humans in that process. And just to, I think James was right. His newsroom analogy, I think, is really interesting. So at WMG, we, our partners are many of the world's largest, most trusted news organisations. We have Wall Street Journal and the Washington Post, etc. All of them are using AI in the newsroom. All of them are crunching data with with AI. They're all doing a lot of analysis with AI. None of them are replacing journalists with AI because you still need a human touch to work out, oh, what is it about this story that's really interesting? What's the angle here that's going to resonate with my reader? What's the next question I want to ask to take this story further? And I actually think taking that that attitude or that approach in the marketing room, like a newsroom, it's kind of an interesting metaphor. I think that works quite well. I

Ben Walker  24:44  
thought I thought that was a great metaphor. It appears in the reports as well, and it's what is an outstanding analogy. But there is a risk there. There is a risk there. You go too far, and we get this similarity. You know, this this this this the the thing that we're looking for. As brands, as marketers, is differentation. We're losing that very thing that we seek. And Gemma, in the report itself, there's some interesting nuggets about that. Is that the risks that companies perceive to going too far with this technology?

Jemma Conner  25:16  
Yes. So AI is something that people pretty universally are wary of, and there is a wealth of Hugo polling data to evidence that. And from the last few years, it's something we've been really focused on. But for this particular project, for marketing decision makers as a whole, after data privacy and kind of beyond that initial issue, the biggest dilemma that people say they've experienced or they expect to experience in the future is AI producing inaccurate or misleading data? That concern is fairly uniform across the sample, regardless of company size. With around a quarter saying that this is the biggest risk facing them. However, if we dig into that data a little bit further, there is a slightly different story. So, if we start to look at responses by the size of the respondent's company, while inaccurate data is the biggest concern for those working at smaller, medium-sized companies, those working for larger companies are more likely to be concerned about marketing becoming too similar across brands with the use of AI. With 30-5% of that group saying that this is the biggest risk they experience, they have experienced, or they anticipate facing in the future.

Ben Walker  26:24  
Well, corporate risk, James Delves managing corporate risk. There's a tension in the market, isn't there, between the desire to accelerate with this technology and actually manage the risk of using it. Spend all this time with the IBM Ryan tables. How do we balance it? How do we balance the freedom to innovate with managing risk,

James Delves  26:42  
and it's a big thing which everyone is basically slowly come to terms with and trying to work out what it means to them. The thing you need to really consider is how do you manage the risk without killing your team's pace. So one of the big takeaways from the IBM roundtable was that innovation without government isn't agility; it's liability. It's not only a real challenge for big multinational teams either. Leaner small SME teams stand to gain hugely by not going with the big LMM brands, but picking smaller task-specific AI models that give them tighter control. They lower the risk. They close alignment to the brand's voice by focusing on picking the actual AI tool that actually meets their needs, rather than just the big shiny one, that may work much better for that kind of medium-sized or small enterprise. But leaders need to be really clear and need to have frameworks in place that protect their IP, secure the data privacy while still giving their teams the kind of sandbox to play in to experiment safely to work at how these tools work. But going back to Jamie's point earlier is that you know 69% of consumers hold the brands entirely accountable for AI errors. 33% blame the AI software providers, if there was an issue, so basically the burden is on us, is on marketers. So we need to use transparency like a core signal. It's not about ethics; it's about almost a commercial necessity. So it's really putting those guardrails in place, getting that governance really tight, and then letting people still be creative and still have the time to play with the stuff, but do it in a way where it's not going to cause, you know, government issues, etc. And I think that's easy in a small organisation, but it's going to be really, really key going forward. And a lot of organisations are trying to work out how they do that, and it's going to be a an an ongoing challenge.

Ben Walker  28:40  
Gemma, is there much evidence? Is there any evidence that consumers trust this stuff much at all?

Jemma Conner  28:46  
Not really. No. So we've done lots of polling on AI in in various industries at YouGov as part of our commitment to public data, and people are very wary of it. Not just in the UK, but with my my European hat on, it's it's a real issue beyond the UK borders as well. People are very very cautious about how it should be used in terms of how it's used in in work and in across different industries. The line seems to be kind of decision making and when ethics come into it. So they're quite happy for AI to be used for administrative tasks to speed up processes, to kind of take the faff out of your day-to-day life at work. But the when it comes to actually making decisions and something that might actually impact people's lives, then they want a human to step in. So in 2023, we did a white paper about AI across education, police, and or the judicial system and medicine. And people were very clear that it was fine for kind of AI to be used to book GP appointments, but they wanted to see a doctor. Juries should be people rather than AI. There's a real limit to what people are willing to put up with in terms of what AI is used for, and we see this in the CIM polling as well. People hold the the brand responsible if there's a an error or harmful content in AI-generated advertising is the brand that is responsible, rather than the developer. Seven in 10 say that it's the brand that that should hold responsibility for that mistake, compared to just a third saying that it's the AI developer. And there's a real clear view from the public that this is not regulated enough, and particularly through the CIM lens, AI-driven advertising should be more tightly regulated. Is very clear coming through this poll, and 73% say the regulation currently isn't enough, and that's in line with other use of polling. The public want to see regulation over AI innovation.

Ben Walker  31:05  
We'll dig into the regulation aspect a little bit, a little bit later. About personal data, we're a little bit of consumer data, personal data, don't we, Gemma, in marketing? People happy to give this stuff over to robots?

Jemma Conner  31:18  
Absolutely not. No, 70-8% say that they're uncomfortable with AI using their personal data. So this is really quite an overwhelming finding in polling terms. And again, the younger people are not particularly pleased about this. 80% of that group. The older generations are really, really uncomfortable with this. The over 60-580-8% of that group say they're uncomfortable with AI having access to and using their personal data. So this is not something that people are willing to to give up, and they they want humans involved in that aspect.

Ben Walker  31:52  
Finally, we found something to unite the boomers and the Zoom. It's taken a while, but we found something, Jamie. Um, how are we going to deliver personalised experiences when neither Boomer nor Zoomer, and probably most other generations, don't like giving our robots the data.

Jamie Credland  32:06  
I think it's a it's a really core issue for advertisers in general, and it has them AI has massively accelerated this issue, but I don't think it's a new issue, right? We've all had brands use our data in what feels like a slightly creepy way when you get an email that you know references some piece of your personal information that you didn't expect it to, or something like that. So, from a marketing perspective, yeah, and I understand, I absolutely understand those consumers being very concerned about AI using their data. But I think for a marketer, the rules are kind of the same as they always have been, which is one, be as open and transparent as you possibly can be with your customers about how you're using their data, what you're trying to do, what data you're gathering. Don't gather data that you you know don't need for us just because you can do it doesn't mean you should do it. And if you are going to use that data, use it in a way that actually delivers some value for the customer, not for you as an organisation. So you know, when you go to the supermarket and you get a text on your phone on your way in saying, "Oh, by the way, you always buy this bag of crisps. We're going to give you 10% off that bag of crisps. Lovely. I will. I'll happily take your 10% off that bag of crisps. When you get a slightly creepy message when you're not thinking about the supermarket at all, saying, "We know that you've spent X 100 quid on crisps in the last year, and by the way, here's some diet pills for you. Yeah, that's not a good feeling. It feels awful, and I'm, you know, from Martin's perspective, it might make perfect sense, but from a customer perspective, it's lousy. And I, so I, I think that has probably been true for a while. The other thing that I see as an opportunity area actually is this whole piece around contextual advertising, certainly online on on the web, where contextual basically just means you're you're choosing to place your ads on pages that are relevant for an audience, so you don't need to do complicated things with cookies and demographics of who that audience is. But if I'm reading a website which is the top 10 new loudspeakers for your living room or something, and you want to advertise a loudspeaker thing. That's a perfect place to do it, and AI is enabling advertisers to do that in much, much smaller ways than they were before in terms of understanding the content on the page and what messages might be relevant for that page. So it's still targeting, it's still delivering marketing efficiencies, but it's not creepy. As a user, you're like, "Oh, this is a brand who has found the right moment to talk to me. This is adding value to my experience. Lovely, I'll go and find out more about that brand. And so, you know, that's using AI, but not in a way that I think the consumers are worried about when they're answering that question. I think when they're answering that question of "I don't want AI using my data, they're talking about AI using their personal data, which I completely understand. I think there's ways that marketers can embed AI in the process that gives them gives a personalised experience without being invasive.

Ben Walker  34:52  
Some might say the better regulation might do the answer. You touched on it earlier, Gemma. Do you want to run us through? What you found about attitudes to regulation in marketing vis-a-vis AI in the research?

Jemma Conner  35:06  
Yes. So, given we've discussed such high levels of distrust and discomfort with AI, it's probably no surprise that the public find the current approach to regulating AI-generated content lacklustre. And just 7% say that the current amount of regulation is correct, and compared to 70-3% who think it should be more tightly regulated, as before, 18 to 20-four year olds feel more strongly about this than other groups. With 80% of these age groups, then more regulation is needed.

Ben Walker  35:35  
It's interesting, isn't it? I just got the graphs here and I scribbled Z, M, X, and B on there from Zoom, Millennial, Yennet, and Boomer, and again, it's another one where the youngest Zoomers and the Boomers are most pro-regulation. So again, it's something else that this nervousness about the technology does seem to be something that unites those two generations,

Jemma Conner  35:55  
probably for different reasons. That I think that's what's interesting. It's they're driven by that, that reasoning behind that opinion is coming from completely different places. I think that that older generation, it's a real fear of the data, but not necessarily haven't been exposed to it that much, and that's maybe where the fear is coming from. But then the younger groups, it they have been exposed to it and they don't like what they see.

Ben Walker  36:21  
Interesting, interesting. Interesting. Do we think, Jamie, that regulation is keeping up with innovation?

Jamie Credland  36:28  
No. I think that's that's the easiest question of the afternoon. Fantastic. I think when we talk about AI regulation, it's yeah. For the purpose of this conversation, I think part of what we're talking about is is labelling on ad creatives that have had AI involved, or you know those kind of directly consumer-facing marketing executions. But for the members of WMG, there's extremely little regulation, if any, about these tools scraping the hard work of journalists and creators and feeding them into their systems. So whether you're a major news organisation or whether you're a small singer-songwriter, all of that content and published authors, all of that content is being taken in with no copyright payment. You have severe concerns around mental health, particularly around young people, but in fact, in general, of unregulated lack of regulation on tools like OpenAI, where these tools are then giving mental health advice or other medical advice. This whole space needs some regulatory overview, and in fact, I think marketers, because of their, it's in marketers' interest to create trusted environments and to support trusted environments. I'm I'm optimistic that marketers who tend to self-regulate, right? In the UK, we have a self-regulatory body for the for the most part in the advertising space. I'm pretty confident that advertisers in the UK will get together and say that this is what best practice looks like, and we're going to support it. And I'm sure CIM will be leading the charge there. Unfortunately, we live in a global economy where that might not be the case everywhere, and that lack of regulation-it's-it's concerning. I think it's it's a real concern, and marketers should be concerned about it too. Because if you get to a point where consumers are absolutely distrustful of the messages they are receiving, it makes it much harder for marketers to to talk to people, and you know, it's not in any of our interest to have a distrusted media environment.

Ben Walker  38:26  
It certainly isn't. Just a very good question, though. We we can split hairs over the generational thing. There is a generational aspect to this, but overall, three quarters of people nearly think that AI driven advertising should be more tightly regulated. As Gemma says, it's an overwhelming demand from the public. It does beg a question, though, does it not, James Delve? That within our own organisations, who should own this governance? Is it us as marketeers, legal team? Is it the IT team? Does it go right to the Jamies of this world, the chief executive? You could

James Delves  39:00  
boil it down to two things, really. Part of the conversation is around marketing capability. If you have the skills and you have the discipline, and you produce a really tight marketing campaign that adheres to everything, then you don't have a problem. It's when you don't have those skills, you don't have the guidelines in place, you don't have the infrastructure, all the support, etc. Then it's going to become exceptionally more difficult and risky to basically customers because you're not adhering to traditional marketing principles. A lot of the things which Jamie mentioned around AI, if you look at you know things code of conduct, you just wouldn't do because they're just not correct. But you could roll it back to that. Going back to your question about who governs it, it's really interesting. We've had conversations with HR teams saying we need to own it. We've had conversations with board members saying it needs to be up the board. It's way too big for market. It needs to be at the base of the board. But it goes back down to you know marketers often need no they're the. Of the connection point on a lot of brands to the customer, so they are connected with the customer. They do understand what the customer wants and what they need. And to Jamie's point earlier, this technology is going so fast that putting any legislation in place is really difficult. And you know, I've I've I've followed the AI Regulation Act of government. It's difficult because it is moving so fast, so I think probably it's governance across multiple areas of the business, and everybody needs to have a say in it rather than own it. And once we get that in place, you're probably in the safer space to operate.

Ben Walker  40:34  
Interesting. Let's let's let's look ahead. Let's look a little bit ahead a bit. One thing we do know about this technology, regardless of the advantages and the opportunities and the threats and the risks associated with it, is that it's ever advancing, right? It's it's generally generally getting better. It's becoming more potent. We've even noticed it, haven't we? All on on a micro scale with the likes of ChatGPT. Compared that to three years ago, its accuracy and so on to To its ability now. Nevertheless, that prompts a question, or it does for me anyway, which is: Are there some elements of our business which should never be done by AI, regardless of how good the technology becomes? For example, could we see a difference between whether we deploy it in you know generating advertising versus offering customer support. Jamie, is there some areas that are sacrosanct that always should be commanded by humans and humans alone?

Jamie Credland  41:35  
I hesitate to name a specific area because I think time will not be my friend, and we'll have moved on. But I do think, you know, there is so much evidence. You know, the other trend that is happening in the world of marketing at the same time as AI is this growth of the creator economy, led by individual personalities, and the growth of in-person live events, people meeting, and having brand experiences in their experiential web; those two things are connected, right? And it's because your customers are crying out for a human connection on some level. So, while you might be able to, you know, I've used chatbots on on websites to answer my tech questions with a particular product, and that's fantastic. But if you have a real problem and your product isn't working, to get hold of a human being on a telephone is is gold dust and can leave you with a really valuable brand experience at the end of it. So I'm I'm really cautious to say, oh, that this is the limit. This is what should never be done. But I think every marketer needs to really think through not just okay, what's the what's the cost saving I can do by automating this tomorrow, and what's the customer benefit of keeping this as human or automating it? And and that's complicated. And I think there was a there's a great quote in the report from Matt Burns from Rolls Royce, and he talks about the future team being an orchestrator of talent, which I really like. And this idea that you're going to have people in the team and some form of synthetic bot who are also working together, and it's this unified team. So wonderful metaphor, except we don't know who's in the orchestra yet. The instruments are still being built. We don't know yet what those technologies are. So yes, we should test. We should see what happens if we make all the call centres run through AI, and and see if we can automate that because it could save us a tonne of money, but the minute we start seeing signals that say it's not actually your customer's renewal rate or your lifetime value of a customer is dropping away, you should be very prepared to say, okay, this this is not an AI suitable capacity. We're going to keep this human led.

Ben Walker  43:38  
James Dell, have you done that flash dance with technology where you've got your technology is broken, and the person you want to fix it is anything other than a robot.

James Delves  43:46  
Yes, yes. My recent laptop, I had same issues. I did certain amount of trouble spotting with an AI tool. It was really helpful, but then I had to speak to somebody to say, "Look, as part of the process, it's not quite there. I hit to this point; it doesn't work. Spoke to person. He cleared it up in absolutely no time. So that combination of AI and person worked really well for me. The other thing we also hear about it is when it comes out things like cultural judgement. Then, when you're going out internationally, or you just need to read the room, or you need to basically understand how different cultures react to things differently. That I still think is a there's a really strong case for a human to be involved in that process, because as we know, scraping all the data to Jamie's point earlier probably doesn't give you a view of actually what's happening at this precise moment in time, and you know the various different sort of nuances you need to now to really make your mess in land with the local population,

Ben Walker  44:43  
do you think that there are some areas where we're overestimating AI's capabilities? The implication of my question is yes, but there may be other areas, Jamie, where we're underestimating its capabilities. Do you think we've got the balance right into how much value we're.

Jamie Credland  45:01  
Yeah, absolutely. So I I think James's description at the beginning of front office and back office is actually really helpful. And I know, yeah, marketing is a complicated function. There's a lot of things going on. Front office, back office is probably a simplification, but I think it's useful. I think we are overestimating the impact AI is going to have in the creativity space. Like, of course, it's going to cut 50 different versions of the same ad. Absolutely, it's going to do different versions. It's going to do translation. It's going to do all that stuff. Great, but it's not going to leave lead creative strategy and really thought leadership and content generation. I think that humans are harder to replace in that space, at the quality end, than maybe we think at the moment. On the flip side, you know, and we've only lightly touched on this, but we've all had to do timesheets and invoices and completing spreadsheets, and yeah, so much of what we have thought of for the last 5050 years as part of office work is either going to become instantaneous or basically extremely quick to do. So I think it makes you think differently about what you're really doing that adds value to the business as a marketer. And if if what you're historically have been really good at is reporting, that's not going to be that very that interesting for the business because that's just not going to be a skill people need. It's going to be automated. It's going to be country through that very quickly, if your skill is I can understand a strategy document and then I can I can build off that and then decide where we're going to take this and do something that surprises and delights our customers. Wow, that's going to be super valuable in a world of AI slop and instantly generated content. So I think your this thing about differentiation is going to become really valuable, and loads of the other business stuff. As we're getting there, I think that's. I think we are still struggling to understand how much of that is going to disappear. Because really, you you you shouldn't need to be doing fiddling around with Excel sheets anymore. You shouldn't need to be uploading invoices for a particular month's report or all that stuff, which like it for not for a lot of people. That is big part of our jobs.

Ben Walker  47:10  
Yeah, the prompt. My final question: people then what my my final questions are usually about crystal ball gazing, which is not popular with people because they feel they're going to be held to it. But at least with this one, we're going to limit that gaze, the term of that gaze to 18 months, which in AI world, I suppose, is still quite a long time. With that in mind, Jamie, how do you see an AI native marketing team looking like within the 18 month period? You know, what does it look like if you can gaze into the future 18 months, we come into a well-run, responsible, effective AI human marketing team. What does that look like?

Jamie Credland  47:49  
I mean, I think this is really difficult. Partly because if you look at marketing, and James knows this better than I do. If you look at what marketing teams look like today, they're very, very different at different organisations, right? Depending on whether they're B 2b and B 2c or international or the scale of the business, etc. But I do think you know things we should be aiming for within the next 18 months are that certainly some of the more technology and data led parts of the media buying process are automated. There are huge gains to be had there that everyone in your team understands how these technologies work and has a shared view of how they're going to benefit the business. And maybe that doesn't mean everyone you doing everything all the time means dividing and conquering those tasks. And hopefully, then also a view of what marketers should be doing with any time that they save as a result of those technologies. So I think this is a bit that's maybe missed out, but if you're saving hypothetically an hour a day thanks to using AI tools, what are you going to do with that hour? Are you going to use it to really think outside the box of some new creative idea? Are you going to meet your peers across the industry to have a coffee with them and understand what's happening in their business? Are you going to meet with suppliers and vendors to understand better how your you can use their services? There's lots of other parts of the marketing job that get underserviced at the moment. Everyone is in front of their screens tapping away. So, in an ideal world, I think you'd have a marketing team where we understand what we want to automate, and we're going as fast as we can to automate it, but then we understand what we don't want to automate, and marketing leaders are encouraging their teams to do more of that stuff.

Ben Walker  49:29  
James, did you want to come in there?

James Delves  49:32  
I completely agree. To completely steal some ideas off one of our course directors, he basically put it down into talent trajectory traction tracking. Got to get the people in the right seats, and then give them tools to do it. You've got to basically know where your team's skills need to be in the future and in those gaps. And then when it comes to abstraction, the tools only work if they're actually built into the marketing process you want to do. The tools shouldn't be driving what you do. It should be. You have a marketing process, and the tools enable it. And if you can't show the output per person, you can't defend investment. So you should you shouldn't just be using tools for the sake of it. And if you get that kind of balance right, the whole newsroom approach, I think you can do some really interesting stuff with AI tools. But it goes down to empowering people to use their spare time for creative and interesting things, not just do more stuff, produce more content. I think we really need to avoid that.

Ben Walker  50:30  
I have to say, fascinating show, guys. Thank you very much for joining us today. It's also a fascinating report, and we're going to that's going to be offered to the CIM membership in around about a month's time, so this was a special preview, a little bit of an exclusive preview into that report. Thanks you to Gemma Connor from YouGov. Gemma, fantastic to get your insights on the report today. Jamie Credland, who is CEO of the World Media Group, and of course mr. James Dells, head of PR and external affairs at CIM. Now, quickly before we go, do complete the survey. The link is in the description to help us improve the show. We do actually read every response you send. We value your input, so it's your chance to shape the future of the CIM marketing podcast. Please give us, of course, a rating and review on whichever platform you're using. Thank you very much again, everyone, for joining us today. That's the CIA Marketing Podcast. It's been a great show. We'll see you again very

Karen Barnett  51:39  
soon. Searching for a way to build your organization's marketing capability, CIM's company affiliate programme provides a simple strategic route to a more confident, high-performing department. Get in touch with our experts today to discuss how we can help you develop talent and deliver better commercial results. Head to cim.co.uk and search company affiliate. The contents and views expressed by individuals in this podcast are their own and do not necessarily represent the views of the Chartered Institute of Marketing or the companies they work for.