The AI Advertising Podcast: S1

Episode 1

Generative AI is Reshaping Ad Creation

About This Episode

Mar 19, 2025

Can AI be truly creative, or is it just remixing the past?

In this episode, we dive into how generative AI is revolutionizing ad creation, from automating copy and visuals to reshaping the role of human creatives. Hear how AI is enhancing, not replacing, creative storytelling.

Randy Newman | CEO, Colour

Te’Shawn Dwyer | Manager, Creative Studio, StackAdapt

James Targett | Creative Project Manager, Creative Studio, StackAdapt

00:00

Transcript

Diego Pineda (00:00:00)

In 2024, Coca-Cola released a Christmas promotional video made with AI. Online users called it “Soulless” and “devoid of any actual creativity” among other things. And just a few months before, another AI generated commercial by Toys R Us faced backlash online from creative professionals and users alike.

Whether we like it or not, Generative AI is here. It writes. It designs. And it makes videos. And for marketers, this sounds like a dream come true—or in the case of Cocal-Cola and Toys R Us, maybe a nightmare.

Today, we explore a big question: Can AI really help us be more creative, or are we sacrificing the soul of advertising for speed and efficiency? This isn’t the first time technology has changed creativity, and it won’t be the last. So, what’s happening right now?

Podcast Intro (00:00:59)

Welcome to the AI Advertising Podcast, brought to you by StackAdapt. I’m your host, Diego Pineda. Get ready to dive into AI, Ads, and Aha moments.

Diego Pineda (00:01:12)

In this episode, you’ll hear from two members of StackAdpat’s Creative Studio, Te’Shawn Dwyer and James Targett, as well as Randy Newman, the CEO of Colour, a digital media agency in North America.

Let’s start with the ‘why.’ Why does generative AI matter so much for advertising right now? The answer is simple: creative teams are overwhelmed. Tight deadlines, shrinking budgets, and a demand for constant innovation have made the job harder than ever.

Here’s Te’Shawn Dwyer.

Te’Shawn Dwyer (00:01:42)

We work in an industry where everybody wants everything yesterday, right? So the faster we can cut down on making the creative work in a timely fashion, the happier clients are, brands are.

Diego Pineda (00:01:56)

James Targett explains how the Creative Studio team at StackAdapt has been testing generative AI for copywriting.

James Targett (00:02:03)

When it comes to AI, we typically use a term called AI assisted. And the reason we use that term is because we realize that even though AI can help us creatively, like it can help us achieve a scale of work, that we haven’t been able to before, like we can produce 50 headlines as quickly as we could historically produce one. Like that’s really beneficial, but we always want to have a human involved. And in this case, it might be like curating those best headlines of the 50 and then workshopping them further to add that human element that we were just talking about.

Diego Pineda (00:02:43)

Gen AI tools—like ChatGPT, MidJourney, and Adobe Firefly—are changing how ads get made. They’re not replacing humans but speeding up the repetitive parts of the job. Think headline generation, resizing images, and brainstorming ideas. And that time saved? Well, it can be invested in strategy and storytelling.

So far, so good. But where is generative AI making the biggest impact? Let’s break it down.

Diego Pineda (00:03:15)

Here’s where things get really interesting. AI isn’t just one tool—it’s many tools, and each one is solving different problems for creative teams.

First, there’s copy generation—writing the words that sell your product. As James Targett described it, AI tools like ChatGPT or Jasper can generate dozens of headlines in seconds. But does it work? Is it effective?

James Targett (00:03:35)

Yeah, so I think there’s a lot to test here. We’ve started tracking,like in the case of copy, right? Cause I think copy is one of the easiest AI things to leverage. So we have clients sometimes who supply their own copy for ads. We have a history of writing copy for clients’ ads.

And now we’re starting to get into the world where with some clients, we use AI as a tool, right? This is the AI-assisted generated copy. And so we started tracking core metrics about how all of these performed, CTR in particular. And the last time I checked, actually, the AI-assisted copy was outperforming client copy and, in a lot of cases, outperforming our internal copy as well.

Diego Pineda (00:04:21)

But there’s a catch. Te’Shawn Dwyer explains the importance of a good prompt.

Te’Shawin Dwyer (00:04:25)

If you think about the prompt as a creative brief, then you can’t just be like, hey, generate this ad for this company. It needs to be ‘generate this ad for this company using this brand voice, this brand messaging for this target audience, you know maybe with this creative reference,’ right? You can see how A versus B gives you a stronger output.

James Targett (00:04:49)

To leverage AI effectively or generative AI, you probably have developed a prompt, right? Like you have your own process, which for, you know, different people, like they might see it as like their own intellectual property, right? Like I probably would not tell you exactly what goes into the prompt that we leverage to get the results that we get. But in the sense of testing and finding out what works, we can test different iterations of the prompt that we’ve developed. So it’s like, you know is V2 doing better than V1? What did we change between those two different versions? And so you know over a span of time, you’ll be able to take your prompt and make it a lot more sophisticated than it was in the beginning. And you’ll be able to correlate specific reasons why it was working better or worse than other prompts you’ve developed.

Diego Pineda (00:05:52)

Next up is visuals. AI tools like MidJourney and Adobe Firefly can now generate high-quality images very quickly. James Targett explains the impact of AI on image generation.

James Targett (00:06:06)

I see a lot of potential when it comes to image generation. So I think when you look at an advertising agency’s process, there’s an opportunity to take parts of that process, such as looking for stock photography, and instead of searching for an image, you can just generate it. So you can get very specific about what you want. If you want to run an A/B test on a particular image, let’s say you want to test an image with a single individual versus an image with a couple, you know, typically you would need to find a stock photography that has those different options, but now you can take a single image as a reference and generate the variation that you want. So there’s some really interesting creative opportunities that are starting to emerge.

Diego Pineda (00:06:53)

But the potential is not only in how we create images but also in how we deliver them to our audience. Let’s hear from Randy Newman, who has more than 20 years of experience developing and executing effective strategies in the advertising industry.

Randy Newman (00:07:08)

You know the way AI, from a personalization or a creative optimization standpoint, is helping is you can deliver a message that’s as relevant to any point in the journey as possible. If I’m online and I’m shopping for blue umbrellas only, then I can get ads that you know feature blue umbrellas. Or on rainy days, I can get served those ads. So it It’ll pull from data based on my journey to anticipate when there’s a higher need and then you know give me a creative that speaks to that need. What AI has done from that creative optimization standpoint is essentially introduced what we call a speed to insight, right? Where the data, you know, it’s customer journey data, it’s behavioural data based on how someone is you know consuming information on a website, all that feeds in real time into the algorithm, which then says, this is the ad for this person at this point in time, and here’s what it should say. And it turns that around very quickly.

Diego Pineda (00:08:16)

Video is where things get a little more complicated. Tools like Runway can turn text into video or animate static images. But video is personal, emotional—and some people worry AI makes it feel cold.

Randy Newman (00:08:32)

AI-generated video is still probably in its early, very early stages. So I’ve seen a lot of criticism of it. A lot of the criticism I have seen comes from production companies that you know they probably feel a little bit threatened. Recently, I did see actually a short film that was produced using AI, and it was excellently done. Quality was very high. Every shot looked pretty authentic. You know, there wasn’t much of a story there. I think the tools are getting better, the capabilities are getting wider, it still takes you know the right prompt.

Diego Pineda (00:09:12)

I asked Te’Shawn how to make the most of generative AI for video creation, given its current limitations. Here’s what he had to say.

Te’Shawn Dwyer (00:09:21)

The opportunity where AI and video come together is about taking like a very iterative approach, right? So I think definitely in that kind of ideation phase where you’re storyboarding, coming up with a multitude of ideas, AI is definitely key, right? But I think that it needs to be kind of mixed and matched with human curation.

Diego Pineda (00:09:47)

Finally, AI is helping creative teams automate repetitive tasks, like resizing ads for different platforms. Te’Shawn sums up how powerful this can be.

Te’Shawn Dwyer (00:09:58)

It kind of enables teams to do less with more. So, you know, it’s removing a lot of these restrictions here. So if I had just had a display ad, but I wanted to adapt that to different channels like Meta, like video, like native, I can take that into like a multitude of AI programs. Photoshop, Illustrator, et cetera, et cetera, and have that adapted for different sizes, then leverage things with copy generation to make that for those specific channels, right? So like now, instead of being stuck in one box, I’ve got an entire multi-channel strategy.

Diego Pineda (00:10:31)

All this sounds great, but AI isn’t perfect. It can ‘hallucinate’ facts, generate content that doesn’t match a brand’s voice, or even perpetuate biases. And without human oversight, things can go wrong.

James Targett (00:10:46)

You know, I do think it’s a good best practice to always review specifically like facts and figures that an AI will produce in copy generations. Because it does, and I think, I think most people know this, it does have the tendency occasionally to hallucinate because it sees this as being creative, right? Where it’s like you know it’s not distinguishing creativity necessarily from factual elements.

Diego Pineda (00:11:12)

AI can’t replace human judgment—it needs human editors to refine its outputs, guide its creativity, and ensure campaigns feel authentic. Here’s Te’Shawn’s take on the role of humans.

Te’Shawn Dwyer (00:11:25)

I guess for brands and agencies kind of starting on their AI journey is just to remember to act as a curator. So as this tool generates a lot of options for you, you know, still provide that final direction, editing and creation to make sure like all these things that are being output match the brand tone, that are brand safe, right? Like you never just want to have it generate something and send it out and then you find that it’s not brand safe at all, right?

Diego Pineda (00:11:56)

What does the future look like? AI tools will only get better. They’ll offer deeper personalization, faster testing, and smarter insights. But the role of the creative professional isn’t going anywhere.

Randy Newman (00:12:10)

So just with that notion that marketing is you know at its core creative, the creativity pieces is something that I think AI is gonna take longer to master, right? Because the way Gen AI works is it looks at everything is being created and it kind of almost gives you an average, in record time, so that’s why it’s efficient and effective. But at the same time, it’s pulling from you know all the data that already exists.

Where as sometimes from a campaign strategy perspective or even from a targeting perspective, what’s needed is innovation or to do something that hasn’t been done yet. Given what limited my understanding of how LLM works, it’s not going to come up with something that’s never been done. It’s going to summarize everything that has been done.

Diego Pineda (00:13:04)

And for marketers just starting out with AI, the advice is simple.

Te’Shawn Dwyer (00:13:08)

Start small. There’s no reason to sort of be like, hey, I need to generate 50 creatives, dump it into this program and do it. So starting small with AI and trying to automate tasks that are just kind of really redundant. So like maybe that’s generating several pieces of headlines and copy. Maybe it’s image generation. Like maybe you do a lot of things with stock photos. So I think that’s a way to start small and then expand as they kind of get more comfortable. But then also focus on the data. So see what performs best. Test V1 against V2, against V3. Improve your prompts slowly, right? So you know if you find, just as James was saying, that you know maybe you need to include more of the brand voice tone and those things in your prompts and sort of improve it over time. Yeah, don’t forget that, but I would also just add last, just do not forget human creativity, right? So AI is not here to replace anything, but use it as a tool to enhance your workflow and not the process.

Diego Pineda (00:14:11)

Generative AI is reshaping advertising—but not in the way you might think. It’s not about replacing creatives. It’s about helping them do more, faster, and better. It’s about freeing up time for the big ideas and bold strategies that machines can’t replicate.

The key takeaway? Balance. AI handles the heavy lifting, but humans provide the judgment, strategy, and emotional connection that make campaigns truly resonate.

Podcast Outro (00:14:40)

Thanks for listening to this episode of The AI Advertising Podcast. This podcast is produced by StackAdapt. Visit us at stackadpat.com for more information about using AI in your advertising campaigns. If you liked what you heard, remember to subscribe, and we’ll see you next time.