AI's Impact on Marketing and Advertising

“The entire business model on which people have planned their futures is wobbling underneath them right now, and they’re going to have to hang on to that wobbly platform and find themselves a ladder somewhere.”

– Tim Burrowes

What you will learn

• Exploring the impact of AI on media and marketing
• Dhallenges faced by journalists in the age of AI
• The transformation of creative agencies through AI
• AI’s role in enhancing investigative journalism
• Future training and development for young creatives
• Business model disruptions caused by Generative AI
• The balance between human creativity and AI automation

Episode Resources

• Artificial intelligence (AI)
• Generative AI
• Programmatic advertising
• Motley Fool
• Martin Sorrell
• Mad Fest
• Investigative journalism
• Performance advertising

Book

• Media Unmade: Australian Media’s Most Disruptive Decade by Tim Burrowes

Transcript

Ross Dawson: It is awesome to have you on the show.

Tim Burrowes: Ross, it’s been far too long. It’s been a while. It’s been a pandemic since we last spoke.

Ross: Oh, yes, the world has changed and continues to change as we speak. 

Tim: It certainly has, I reckon the last time we spoke, the world was still talking about the possibilities and excitement of AI when it finally arrived one day.

Ross: So you have been central in the world of media and marketing. And as you say, the people talking about AI and edge cases and programmatic advertising and a few kinds of very focused things. But now AI has arrived. How does that change media and marketing in three words or less?

Tim: In every way?

Ross: Got it.

Tim: The truth of it is different things for media, different things for marketing? Gosh, I tried to find where I can, the case is for optimism and positivity. And I guess, in the same way that horses and carts gave way to a thriving automobile industry. And it feels a bit like we might be at that stage for the media and certainly for communications agencies where they’ve got such big disruption coming along. And, of course, so many possibilities and new ways of doing things. But it feels like the entire business model on which people have planned their futures is wobbling underneath them right now. And they’re going to have to hang on to that wobbly platform and find themselves a ladder somewhere. Because, you know, obviously, there’ll be a way through to the other side, because there always is, but wow, we’ve never seen change, like..

Ross: Yeah, well, arguably, you know, major magazines have been pretty wobbly for a long time like this. There’s no single year which hasn’t had its damage. 

Tim: Yeah. I mean, absolutely. I mean, I’ve been a journalist since 1989. And the theme when I walked into my very first newsroom, well, firstly, and trained on a typewriter, manual typewriter for the first few months, but was it was just as the printing of the newspapers was digitized, a whole bunch of printers were in, in the process of being made redundant right then. So yes, we had this weird kind of battle of the humans against the computerization where, as the sort of protest these these printers, who knew they were doomed, but we’re still at this stage, laying out the newspaper each day, with just put subtle sabotage in while they were having they’re kind of they could see what was coming down the track. So you’d have to be very, very careful because things like the not in not guilty would get removed in articles, and he would become che and all of these subtle things, which were quite hard to spot on the final round of proofreading as the as as people went, when kicking and screaming into the night, and I, that was the printers and sadly, I think it might be the turn of some of the journalists and,

Ross: Oh, let’s look into sort of media and marketing. But I mean, company, the theme of amplifying cognition. Alright, so journalists are super smart. And I’ve always said, you know, if you’ve got a journalistic training, you can do well in the world, because you’re able to pull together information makes sense that will communicate well, you know, these are fundamental skills and will continue to be, but how can you know, good journalists today? Use AI? Or what is their relationship to AI? I mean, obviously, there’s going to be a lot of AI reporting, but what are the complementary roles of good journalists in AI today? Oh, what could it be?

Tim: There’s no one answer. Obviously, there’s several great examples. And I suppose the one that’s given me the most pause when I know, because if you want to think as a journalist, okay, things are going to be fine. I’m smart. I know my bait in a way that no, I have occurred, you know, I, you know, I’ve been writing about media and marketing for more than two decades now. So, so my, you know, my, my neural pathways are trained and possibly calcified, to see the world through the eyes of marketers, I suppose, because they’re my constituency. So, you know, in my time In Asia, you know, I’ve, I keep telling myself, you know, I, you know, I’m specialized enough that I can probably go fine. But then I had real pause when. And we might talk about this more in a minute. But um, we created effectively a chatbot called timbre, which was based on a book I wrote a couple of years back called Media unmade. And then based on the content of the unmade newsletter, since we started that, and I just remember asking it, and it was an example of showing somebody something about the outdoor advertising industry. And the answer it gave, which obviously, had been based on my writing. But what really struck me was it, it got all of the obvious points I would have made, and then it made an extra connection about an outdoor company, I’d obviously written about, at some point, a small one, because it was featured, but I had forgotten all about. And it, it, it answered it just despite the fact that it’s based entirely on my writing. So it’s only training from that. It got there slightly better than me. So that kind of gave me pause a little bit. Um, so I think it doesn’t entirely answer your question. Because you know, that’s an example where it goes above and beyond, because, but yes, I guess in my super specialized niche, hey, yeah, it’s really useful. Having a really specialized reference tool, trained on everything I’ve written, I can ask questions often. Give me that information.

Ross: Just it’s just on that point, though. In that case, you were able to judge that what the machine created then was insightful. Yes, he doesn’t generate a whole bunch of stuff. And there was probably another point of me, which wasn’t very, which was boring, the new, new, new, which was insightful. And that’s so that’s, again, you know, there is a cycle where you can feed the machine, what comes out of the machine you can then build on? 

Tim: I think that’s a very fair point. And I think, of course, the other thing is, you see the possibilities for journalism. So for instance, something that I’ve been researching as we speak, and I think it probably by the time this podcast goes up, will probably have already been published is that I’ve been looking, looking at the finances of an Industry Foundation. And they’ve published via the charities commission, they’ve they’ve, there’s six or seven years worth of their data. And my instinct as a journalist is we’re almost at the point now that rather than me having to laboriously open up each one of the balance sheets and manually put in the data and manually draw myself the kind of graphical representations, I can ask AI to do it. But what I noticed was when the moment came today, or over the last couple of days I was working on this, I realized I don’t trust it yet to get it. Right. And I think that’s the thing. We were still at that point. But for journalism. I’m not sure it would have shortened my journey, my time spent on that particular piece of work now, over time, I think it probably will, you know, I think quite quickly will be at the point where transactional stuff like, you know, writing about a company? Well, I think we’re all there already there actually writes from that company’s quarterly results or something that comes through, you can, you know, and the Ansel will sort of be spat out, you know, I kind of think that a lot of what we see on, you know, the basic stuff that comes out and Motley Fool and places like that, I’ll be surprised if that isn’t an element of AI. So, there are things like that. But then, of course, I guess the question is, how helpful is it if it’s the transactional stuff that’s happening?

Ross: So I think, you know, I always go, when judges come out, I mean, the first distinction I kind of made is intent. You know, humans have the intent. And so you thought, that’s a foundation, which I would actually like to look into. And, you know, possibly at some point, the AI can surface that. I think, you know, it’s again, that sort of thing is interesting. So that’s, I think, a critical value of humans. And then you can say, well, if you can get there quicker, it’s far better and picking up the emergent trends. And that in stacks of data is definitely one of its strongest use cases. Yeah,

Tim: Look, I think that’s a great point, because, of course, um, where the intent came from was, what the journalist does is they need to have relationships and sources and it took having a drink with somebody and then saying, by the way, you should look at, you know, so because you need a starting point in the first place. So, yeah, you know, I definitely think that those tools will become increasingly useful.

Ross: I think it’s and I think one of the, if you look at it from the positive perspective, you know, investigative journalism is fundamental, it’s critical. And that’s the biggest downside of, you know, lower budgets for newsrooms and journalists. But if the best investment of journalists is to have access to AI, where they can more readily surface and dig into stuff, then that’s going to be good. Yeah,

Tim: I think so. And I suppose the point is, yes, we’ve got this crucial need for society for investigative journalism. And I, you know, I think one of the, the big worries for me is just the side effect of the rise of generative AI is if, as a user, as a member of the public, you can just ask a question of your, of whichever chat bot you happen to be talking to, and it gives you the answer without sending you to the website, then that breaks the entire business model of media. And there’s no, there’s no guarantee that a new business model emerges. So of course, something happens that potentially stops us from funding all of that, investigative journalism or certainly at the, you know, the current level. So it becomes another disruption along that kind of line of disruption. We’ve been Yeah, we’ve been saying for at least the last 30 years, if not more. Yeah, the

Ross: Yeah. The business models of media are a tricky question. And it’s, but yeah, I think that there’s whatever resources we have, we have to just be able to amplify those as much as possible. And, you know, not just churning out, or just a courier reviews of quarterly results, or soccer matches, which can be done pretty decently with.

Tim: Yeah, I mean, I think this is my issue at the moment is we have this sort of strange situation. So the little niche, I’m sort of in the trade press writing about the Media Marketing World, there are five or six titles. And sometimes if the same press release comes in, you’ll see. I mean, we don’t, we don’t tend to do press releases. But you’ll see in most other places that a version of that same story happens. Now, what’s actually beginning to happen already is a PR person is using the assistance of a chatbot, or chat GPT or whatever, to create their press release in the first place. It’s then arriving in the newsrooms now one of those titles, has already quite openly said for all of those transactional news stories, then are feeding them into AI to create a separate news story, which is almost a point of difference for them at the moment. But when every single one of those outlets is doing that we’ve got this bizarre system where AI is talking to AI and there they become a series of filters before they maybe just maybe reach a human. And I yeah, this particular publication, I think it’s hugely Orwellian. They call this practice fast news, which just actually sent a shudder down my spine.

Ross: Yeah, while loop which, you know, again, goes back to the human role, which is where do we put the humans in the loop? And all this process? And so it’s, there’s going to be humans in the loop. We pray, and in which case, so you’re working out where those are. But let’s let’s turn to marketing agencies, that have typically been profligate paid high salaries, and being all these masters of the universe, in them, and now, the lot of them are looking up and saying, oh, suddenly AI is doing things which are creative, and we were where we’re supposed to be the greatest people. We’ve got creative directors. We’ve got executive creative directors, who get paid a lot for being creative and directing. And so what’s the high level view of what happens to agencies, creative all I argued, well, supposedly creative agencies in the world of AI.

Tim: I think one of the big challenges for creative agencies and again, this certainly predates the rise of generative AI but was already on track is, so much of the business model of agencies has been. They haven’t actually ever really been paid on the thing their best, which is the idea is almost always the idea is given away for free. And then the way they make their money is by charging the client for the production work and the hours and all of the things that go around it, that all of a sudden generative AI can do. Now, in the short term, I think for some agencies, that’s even beneficial. You know, there are, there are agency bosses who talk to who aren’t telling their clients how much they’re already outsourcing, but charging head hours for. But that’s a very temporary situation. You know, there are smart procurement people working with all of these big companies who next time they run a tender. And generally, you know, an agency might be agency of record for perhaps three or four years before the lease, there’s a review, if not a full pitch, then those clients will have a really good understanding of what actually can be either taken in house by the client, or just expected to no longer be done by humans. So that that, that potentially breaks the business model in the same way it breaks the business model for media, which isn’t to say that creative agencies don’t have a huge potential to be the contributors to a brand’s business, you know, there are, there’s so much evidence that the right big creative idea helps people sell more cornflakes, or whatever it is. And that becomes even more important, and arguably, for a really big landscape of agencies, and probably too many agencies, the ones that will differentiate will be the ones who can still nail the big idea, not the ones who are kind of in that muddy middle that can be done by AI. But again, it just becomes the question of how do they get to that point where they’re rewarded for that big idea that only a human could come up with?

Ross: Yeah, person structures. Now that applies across other professional services as well. And I think it’s really interesting, looking broadly at professional services of all kinds, and the roles of humans plus AI and service delivery, and then, you know, relationships and client value, and then the similarities and differences between the professional services. But this, you know, this does come back to a point where, you know, we were discussing earlier around, you know, getting the, the young people exposed to the learnings and the growth, which enabled them to get to where they’re able to create, yeah, be the genius who can shave that big idea? 

Tim: I think that is exactly the issue is, you can sort of think now, okay, you’ve got these really experienced executive creative directors who have a great idea. And then they can use these tools to make it and that’s great, you know, for the Agency for them. But the question is, what happens when those people have retired, and the next level of big brain creatives haven’t learned on the job learned on the tools are all of those things, which they, they just learned by osmosis, that gave them the ability to have the big idea. At the same time, I suppose I was chatting, I think back to my experience of perhaps 10 or 15 years ago, just going on a very basic video editing course. And I, you know, we, this is where I worked previously, when I worked at Mumbrella. And we used to put out a fair bit of video. And I never actually, in the day to day ended up doing any video editing myself, but it gave me a language and an understanding of things I could then ask for, you know, it’s just things like, you know, I remember there was a video I was so pleased with, and it was just the really sounds so obvious when you say it, but for the, for that particular piece, just to edit the piece we were doing to the beat. Now, that’s just such a really obvious thing to say, once you’ve, you’ve been through the process, but I would never have thought to suggest it if I hadn’t got that language. And the thing is, now, you just have your piece of video and you throw it into there’s myriad tools, and it would spit it out at the other end. And you probably would never just have that little idea or that little piece of direction. So I think that is the worry that that kind of the you know, if we don’t if the Yeah, if AI saves us from needing to work on the building blocks, then are we ever gonna be in the position to put the roof on?

Ross: Well, I think that applies. Yeah, that’s absolutely not just for agencies anymore. It’s across all professional services, arguably, you know, almost any, any industry or any work at all. But going just coming back to this frame of Yeah, so humans plus AI so you’ve got let’s say you got your super creative director or if you’ve got some pretty good one. So what specifically do you think of the roles in AI? Good can support these people to do their work better.

Tim: And I suppose, in part, you want to think about the different parts of that communication agency chain, because it feels like for instance, when it comes to, at the end of the story after the the ads have been created, the the planning and buying of the media is already moving in that direction of automation. Anyway, you know, obviously, we’ve seen the rise of programmatic advertising, which is starting to look a lot like a net negative as it happens, because an awful lot of advertising spend is disappearing somewhere along the way. But certainly the ability to plan and target in a really sophisticated way who the audience is, that becomes something which I could do so much better than any kind of old school planter, if it’s done well. And if it’s done in a kind of non fraudulent way. But yeah, it certainly feels that Yeah. That means that an awful lot of the manual labor part of planning and buying media is gone. I saw Martin Sorrell speaker mad Fest in London last year, about this time last year, actually, the way he just put it was he just listed the consequences just to throw a line, throw away a quarter of a million media planner jobs gone around the world, obviously. And and, and I think he’s probably not wrong. On the creative side, so much is, you know, it’s, it’s, every element of production becomes cheaper. Now, obviously, there’s been a lot of talk about what for instance, saw is coming through what that’s going to be able to do when it comes to creating films and a prompt was playing with a tool just today, which probably wipes out jingle makers, you know, as an advertising jingle makers, where you can literally just say, sad country song about a chocolate bar, and it will churn something out for you. It might be terrible, it might not. But the point is, by its very nature, most advertising is average, and half of it is below average. And that’s the stuff that’s made by humans. And that’s the stuff that is very replaceable. So it feels like the, the, the part that’s easy to say, okay, I can say on that stick around as the the big budget, Big Idea stuff, but it’s all that transactional stuff, the, you need to hit somebody with 10 subtly different variants of the same or have similar phone package, but you’re targeting slightly different demographics or different price points or something, yes. And at the click of a button, you can create every version of our ad. All of those things that you know will make the production of the stream so much slicker. But of course, in the very process challenged the business model of making them in the first place as well.

Ross: So we’ve got the AI to complement the greater people. So this, this goes a little to the idea of how can I, you know, I suppose the demand elasticity? So is there a fixed amount of demand for quality marketing? And if so, then, you know, if it’s a fixed amount, then certainly a lot of work that is done currently can be replaced by AI. But if there is more demand, if there is more potential to create more or more quality, then that’s the potential for the future is, you know, do we have an unlimited amount of demand for quality, marketing and to your point, I mean, part of that is then personalization where you have a big idea and then you personalize that idea to every person on the planet or whatever it may be. And so that this applies across all work is how limited or unlimited is the demand for this work?

Tim: Well, look, it’s a fascinating question. Let’s I guess one of the orthodoxies of advertising is you get your brand building, which is kind of slightly nebulous, slightly hard to measure, but you also have your performance side of sales, you know, save 20% Do your sale now and you know, how many you gonna sell tomorrow or you’ll know very quickly with your campaigns working. If we are just just for the point of this debate, just as, you know, marketers would shudder, well, let’s just park the brand side for a moment, then if you sort of assume that with the performance side of advertising, you can measure ROI. And one of the things about this is you should be able to measure it quite well. Then of course, there could be an infinite demand or nearly infinite, if you can tell that every dollar you sell makes you more than $1 in profit at the other end for the extra products that you sell. So, so for that, that side of you, yes, there there can be a demand i i suppose the the question is that sustainable in the long term, when, in 510 20 years time, the wise old practitioners are gradually retiring. Is there actually anybody left who’s got the training to work with those tools that that part or, or more to the point, if this is going on in every single industry, then we may have a lot of very?

Ross: Yeah, well, that’s one of the macro questions, of course, is, you know, do we do good. Yeah. Though, all of this talk of amplifying, you know, increasing economic activity X over my years, is predicated on the fact that there is no distribution and jobs. I mean, personally, I tend to believe that there, we are more likely to have a happier scenario than a dire scenario. But, you know, that’s kind of, you know, that’s a long and different debate. But I’m sure you have been using the tools and all sorts of ways in your own work and I just love to hear about any. You know, what, what do you find AI useful for in your work in your life? How are the insights you’ve gained on what’s useful and practical and makes you able to do More?

Tim: Yep

Ross: That’s really interesting, you know, all these points around augmenting, you know, doing what you’re doing, but then sort of making it richer and more engaging and you’re able to use more so. So just to round out I mean, any prognostications on any of the sort of things we’ve been talking about Yeah, well, it’s a little while ago, my niece, or your granny said, was embarking on a journalism career at school. She said, Okay, we’re going to do journalism as you know, who I’m not sure about that. But anyway, she’s been extraordinarily successful. So lots of myself that we can still be you and I can sort of say, Oh, I don’t know about that. But we’ll still carve wonderful careers forward and I still I just believe in journalism, I believe in what it does, I believe in what it creates, and yes, the business models of fraud and but you know, I’m, I’m an optimist. And so let’s hold a glimmer of hope.

Tim: Yes.

Ross: Thank you so much for your time and your insights. Tim. It’s been a great conversation.