Quad is a global marketing experience company that gives brands a frictionless way to go to market using an array of innovative, data-driven offerings. With a platform built for integrated execution, Quad helps clients maximize marketing effectiveness across all channels. It ranks among Ad Age’s 25 largest agency companies. For more information, visit quad.com.

Rob Rubin (00:00):

In marketing, everything must work seamlessly or efficiency, speed, and ROI all suffer. That's why Quad is obsessed with making sure your marketing machine runs smoothly with less friction and smarter integration. Better marketing is built on Quad. See how better gets done at www.quad.com/buildbetter.

(00:31):

Hello everyone, and welcome to the Banking and Payment Show, a Behind the Numbers podcast from EMARKETER made possible by Quad. Today is July 8th, 2025. I'm Rob Rubin, head of business development at EMARKETER and your host. Today we're going to talk about how AI will transform bank marketing. I'm really excited to have AI analyst Jacob Bourne and banking analyst Lauren Ashcroft join me for this discussion. Hey guys, how are you today?

Lauren Ashcraft (01:00):

Good, how are you?

Jacob Bourne (01:01):

Pretty good. Excited to be here.

Rob Rubin (01:02):

I wanted to ask you guys, just before we jump into the topic to warm us up, what's the most fun thing you do with AI? In your personal life, playing with it, what do you enjoy? What's the most fun?

Lauren Ashcraft (01:15):

Well, I'm having a baby in October. I don't know if I told you.

Rob Rubin (01:18):

What? I did not know that.

Lauren Ashcraft (01:18):

I'm so sorry that it's coming up on a podcast.

Rob Rubin (01:18):

Congratulations.

Lauren Ashcraft (01:27):

Thank you. I recently gave it characteristics of names that we're interested in and asked it for 10 suggestions. I hated them all, but I thought that was a fun process.

Rob Rubin (01:40):

Okay.

Jacob Bourne (01:40):

Yeah. I don't know if I'd always use the word fun, but I think in general it's nice to be able to just ask AI the most random questions that wouldn't really necessarily be appropriate for a traditional search query and you might not want to ask other people and you can just ask AI.

Rob Rubin (01:58):

I'll take that to my example, which is my dog's name is Buster, and his full name is Mr. Buster Snuggles, and-

Lauren Ashcraft (02:07):

That's adorable.

Rob Rubin (02:08):

Mr. Snuggles if he's misbehaving. So I asked ChatGPT what Buster would look like as a human.

Jacob Bourne (02:18):

That's a good one.

Lauren Ashcraft (02:18):

I like that.

Rob Rubin (02:18):

And it was hilarious. ChatGPT nailed it.

Lauren Ashcraft (02:22):

Oh my gosh. You're going to have to share that with us after.

Rob Rubin (02:26):

It is hilarious. Anyway, let's get right to our first segment today, story by numbers. In story by numbers, I pick a few numbers or a number to help us frame bank marketing. And today I picked a huge number, $42.6 billion, and that is US Financial Services digital ad spend. It's our forecast for 2025. Now, that's not all B2C spending for sure, but 42.6 billion divided by the number of adults in the US is $165 a year for every adult. So they're spending a lot of money already. So let's talk about how AI is getting itself involved because there's a lot of money there. How has AI changed Finserv's approach to advertising or has it really?

Lauren Ashcraft (03:21):

I feel like it's really just kept bank marketers constantly in an evolutionary state because it started off allowing banks to better personalize, better target. And there is some content creation, although it has to be heavily, heavily vetted and edited by humans to stay compliant. For example, agentic AI is starting to happen in the banking industry, and that will change by making, for example, chatbots way more able to completely manage customer situations from beginning to end.

Rob Rubin (04:00):

But will that change the way they advertise?

Lauren Ashcraft (04:02):

I do think there's some possibilities with allowing better understanding of customers through that, which can feed back into data-driven marketing. But what's really about to change bank marketing, I think, it is something that is a challenge, is being able to have your content show up in generative search engines because it's a completely different game than having a great SEO score or showing up in regular search engines. You have to tailor your content to answer questions that people are using generative search engines for. So that's new. And maybe two months from now, there's going to be completely new things that bank marketers have to keep thinking about. So it's just constantly changing. That's how AI is impacting bank marketing right now.

Rob Rubin (04:54):

So right now you're saying, and that's really, I hadn't even thought about it, but all of the SEO people are trying to quickly learn how it's different so that they can become AI experts.

Lauren Ashcraft (05:06):

Yeah, definitely. And the article that we just published, we have a lot of tips in it, but it seems to be really important to be writing content just to answer questions to preempt customers' questions about banking. And then when they search answers for those questions, your bank will come up in the search. So there's completely different tactics that bank marketers aren't necessarily already employing. And who knows what those challenges will be a couple months from now.

Rob Rubin (05:42):

Will they use AI to answer the questions so that AI...

Lauren Ashcraft (05:42):

Yeah, just an endless loop of AI.

Jacob Bourne (05:49):

Well, one good example of that is actually Wells Fargo with its GenAI powered Fargo assistant, that it's actually getting a lot of traction in terms of customer use. I mean, in 2024, it had 245.4 million customer interactions, which I guess was-

Rob Rubin (06:06):

I read that too and that's way over, right?

Jacob Bourne (06:08):

It was double what they thought it would be, which means that people are using it and also I guess having longer conversations than they thought they would.

Rob Rubin (06:16):

Is Erica from Bank of America also GenAI? Because I read that it wasn't.

Jacob Bourne (06:21):

Yeah, I have not heard about a GenAI integration with Erica yet, so that might be-

Rob Rubin (06:28):

Yeah, I don't believe it is, but it's still more popular than Wells Fargo.

Jacob Bourne (06:31):

In terms of use? I mean, it's been around longer, right? I mean, well, the GenAI Fargo is new, so.

Rob Rubin (06:38):

Yeah, no, that's a lot of queries. What about other industries? How are we seeing other industries build AI into their marketing stacks?

Jacob Bourne (06:48):

Yeah, I mean, we're seeing AI all over other industries, especially in marketing and programmatic advertising. Another financial services example with programmatic advertising is MasterCard, which actually saw a 254% increase in click-through rates using AI-powered programmatic advertising. And that's because the AI algorithms can just analyze behavior patterns in real time in matter of seconds, which you-

Rob Rubin (06:49):

And reoptimize.

Jacob Bourne (07:22):

And reoptimize, exactly. And across multiple channels simultaneously. And we're seeing this across the major platforms with Google, Amazon, Meta using AI. Google's Performance Max and Advantage+ got AI enhancements and do a similar thing making these split second-

Rob Rubin (07:41):

And they're doing bidding now. And also the next wave is potentially profit based bidding. So AI, if they had a little understanding of some of the costs and margins of products and when they're advertising, it would be able to also start bidding based on achieving profit goals.

Jacob Bourne (08:02):

Yeah, yeah, interesting. I mean, it seems like there are potentially limitless applications for AI, and we're certainly seeing a lot of experiments. A recent one on the social media front is Reddit that launched its AI insights to read through conversations of users on this platform and create summaries of what people thought about ads or product descriptions. And that actually gave advertisers apparently a 19% boost in click-through rates because people are not just seeing the ads, they're seeing what other users are saying about them.

Lauren Ashcraft (08:37):

Interesting.

Jacob Bourne (08:38):

And AI's allowing that to happen.

Rob Rubin (08:40):

That leads to another area like synthetic customers.

Jacob Bourne (08:44):

Yes, digital twins.

Rob Rubin (08:46):

They're using AI to predict how a campaign might work.

Jacob Bourne (08:49):

That's a huge, huge area, the digital twin simulations where you can have AI generating an entire customer base and also in predicting how they're going to act, how they're going to respond to a particular campaign, and then you of course can reoptimize that campaign before you launch using the data that you get from that simulation. So yeah, it's pretty powerful.

Rob Rubin (09:15):

So it feels like AI is here.

Jacob Bourne (09:18):

It's here. Oh, yes.

Rob Rubin (09:20):

It's definitely getting used in a big way, but at the same time, it's still early days. We haven't scaled the idea of producing marketing content at scale is not today.

Lauren Ashcraft (09:35):

That's what I'm most skeptical about just from the compliance aspect of banking. Just the amount of human oversight needed for content before it goes into an ad is very important.

Rob Rubin (09:46):

But couldn't you have like, as we move to agentic AI, can a bank have a compliance agent that is working with the marketing agent?

Jacob Bourne (09:58):

I mean, that's the dream. That's the dream of agentic AI, is that you have the power of generative AI and then you also add this powerful automation. But I think the thing is, Lauren's right, you still need that oversight. And once you have that oversight, it's still not fully automated. So it's incremental improvements and we're seeing it happen, but I think we're a long way, long, if we ever do get to lack of human oversight completely, I'm not sure, but certainly there's still quite a bit that's needed.

Rob Rubin (10:27):

If you had oversight, if human oversight led to a 1% error rate. So we have 1% of errors because we have human oversight, which is, let's say that's good. Right now, if AI can beat that, why would we have human oversight? In other words, if AI can do it as well, same error rate or better, why would we want to pay a human to do it?

Jacob Bourne (10:52):

And that's a very compelling argument. I don't think we're there yet. I'm not sure.

Rob Rubin (10:57):

No, but you said never. I'm challenging the never.

Jacob Bourne (11:00):

Well, that's a powerful data point, but there's always the question as to whether the AI fully understands human context. And so while that data point might be true in a way, it could be that you always want a human in the loop because people understand people on a fundamental level that AI might not ever.

Lauren Ashcraft (11:21):

That's true.

Rob Rubin (11:21):

Now, let's take it a step further in our final segment for argument's sake, where we will argue nicely, and I can feel the temperature already up there, so I'm happy about it, about how AI is going to actually disrupt and Finserv marketing. So what I've done is I've created five potential areas that we've been chatting about of marketing, where AI is going to be disruptive. And what I want us to think about is which area will AI take over soonest? So what we're going to do, and we've already done this, I'll explain it to everybody listening, is I'm going to read these five out. We've all ranked separately, but not shared what we think is going to be the most impactful to least impactful in the 2026 and then re-rank them again and what we think for 2029 so we can see how we think it might evolve and where we're different.

(12:17):

So the first one is bundling products dynamically. So being able to analyze transaction history, their demographics, their life events to be able to dynamically suggest product bundles and that those offers evolve with their needs as their needs change. So that's the first one. The second is predicting churn, which we kind of talked about. Models trained on behavioral data to create early signals of any kind of dissatisfaction or churn, because retaining a customer is a lot more efficient than acquiring a new customer. So trying to keep churn down. Generating marketing content at scale, something we've talked about. So AI to produce compliant, brand-aligned marketing content like ads, emails, landing pages, really fast and much lower cost than what we have today. Testing marketing campaigns. So AI simulating campaign outcomes through hundreds of variations to optimize creatives and channels before launch, and then helping prospects choose products. So enabling prospects to explore banking products in a chat, and the bots can gather information, pre-qualify leads, and then even pass them to human agents at the right moment. So Lauren, why don't you start with your list? What's your number one in 2026?

Lauren Ashcraft (13:45):

I kind of combined bundling products dynamically and helping prospects choose products because we-

Rob Rubin (13:51):

That's what I did too.

Lauren Ashcraft (13:52):

Yeah. Oh, awesome. Great minds think alike. I ranked them first because we-

Rob Rubin (13:59):

In 2026?

Lauren Ashcraft (14:00):

Yeah.

Rob Rubin (14:00):

Okay.

Lauren Ashcraft (14:01):

Just because some of that's already happening, especially in some more powerful financial apps, fintech apps, and I think it can only get even better from there.

Rob Rubin (14:13):

Jacob, were those your firsts?

Jacob Bourne (14:15):

Well.

Rob Rubin (14:15):

They were not mine.

Jacob Bourne (14:16):

I do have helping prospects choose products as the number one, but actually for 2026, I put bundling products dynamically as number four, not because I don't think it's impactful, but because I don't think the technology is quite there to be able to do that effectively all the time compared to a human. Just in terms of what I've heard about AI analyzing financial data, it can be hit or miss at times. And so I think while very impactful, it's kind of a bit down the road on that one.

Rob Rubin (14:47):

So that's sort of where I went. Whereas one, I put them together both in '26 and '29, I put bundling and chat sort of together, but I put them as the least impactful for 2026 compared to predicting churn as the most impactful. I think that it's the easiest one for them to do, to do a better job finding signals for churn and then producing marketing content. And the difference is that I don't think that they're producing it yet when I say at scale, but I have to believe that they're running whatever they edited, they're asking some AI to clean it up so they're shortening the time that it takes them to generate content at a minimum.

Jacob Bourne (15:39):

Which in itself is usually impactful. So I actually put that at my number two because I think the technology is there despite-

Rob Rubin (15:46):

Yeah, I had it number two, as well.

Jacob Bourne (15:47):

Despite in human oversight still, it's saving a lot of time and money.

Rob Rubin (15:52):

Lauren, which were your least impactful for this year then? Was it synthetic customers or what?

Lauren Ashcraft (15:57):

No, actually I had testing marketing campaigns as number three.

Rob Rubin (16:01):

So did I.

Lauren Ashcraft (16:02):

Predicting churn, I had number four. I want to see more of this. I actually just wrote an article about how banks can, for example, predict that their senior customers are facing cognitive decline with a couple of different patterns for their buying. I would love to see much more of that because there's a lot of opportunity to build stronger relationships with customers and their families and loved ones, especially with things like that. And then generative marketing is just of my personal thing. I'm most skeptical of humans being taken out of that process soon, but I think by 2029, I bumped it way up in my list because I do think that ChatGPT has already improved so much over the last few versions that I actually have no idea what it'll look like by 2029, but by then maybe it will stop hallucinating as much.

Rob Rubin (17:03):

So in 2029, I had the being able to help customers pick products and bundling as one and two. By then, the technology's going to be there. People are going to trust the recommendations more than they might trust them today. And then we have a generation that's growing up with it and they'll be more ready to buy and use AI. I think I read an article that you wrote about how Gen Z are really interested in self-serve.

Lauren Ashcraft (17:33):

Yes.

Rob Rubin (17:33):

So the idea that they'll be able to buy without having to talk to a human might be appealing. And then even though I put predicting churn as number one in 2026, out of this list of five, I put it as the least impactful in 2029, just because I thought it was sort of one of the easier things to accomplish today, and it'll be table stakes.

Lauren Ashcraft (17:57):

That makes sense.

Jacob Bourne (17:58):

I actually put predicting churn in number five for both lists. And the reason why is because you can have an AI that can do that, sure. But it's ultimately up to the company's policies about whether they're going to take action. And I think you do need a human to take action on that. And I think so often we don't see companies taking action on that point.

Rob Rubin (18:18):

I think that's right.

Lauren Ashcraft (18:19):

Yep.

Rob Rubin (18:20):

I think that's like a structural challenge just having worked with banks is that the people who would take action on it would be the local branch people. That's how a bank would think about it. And the local branch people are not incented for that. They're incented to sell products, right? That's just not their incentive. So it's hard to get them to do that. And for a lot of banks, traditionally, churn's been pretty low on the scale of other businesses, so it's not necessarily been one of their biggest issues. But as customer acquisition gets more and more expensive, as the tentacles of technology get deeper and deeper into people's lives, it's harder to switch. So if you have someone that's leaving, you ought to be able to help them not switch.

Jacob Bourne (19:07):

Yeah.

Lauren Ashcraft (19:09):

If it were a ranking of what I think is most important, that would definitely be towards the top of my ranking, just because the tools are there and it is so important, especially as, I mean Klarna is becoming more of a bank. All these huge fintechs are inching more into banking territory and competitions just constantly heating up. If you want-

Rob Rubin (19:33):

Didn't I read in one of your articles that... I'm plugging you and we're going to put all these articles in the show notes, how's that? But that fintechs have more personal loans than banks or credit unions in terms of the percentage of personal loans that are out there. They're owned by fintechs. They own that space.

Lauren Ashcraft (19:52):

And ironically, if banks use AI to enable their customer facing staff to be the knowledgeable, empathetic humans that they are, that's the differentiator right there. So I think predicting churn and having a really nice human that knows you at your local branch reach out and touch base, I think that would be an important strategy.

Jacob Bourne (20:20):

I mean, I think, Lauren, that's a great example of just using AI, maximizing the strengths of AI, and then also keeping the humans in the loop and maximizing the strengths of the human. I think that's what we'll use is let AI do what it's good at and keep humans with the relationship facing end of the job.

Lauren Ashcraft (20:39):

Definitely.

Rob Rubin (20:40):

That seems like a great place for us to try to just sum up where we are with the list. And I think that in terms of helping consumers choose products and even bundling those products together, we all agreed that that's super impactful. And I think where we were a little bit different is whether we think the technology is quite there yet and the consumer trust in using that purely AI advice to act on. And then the other one we talked a lot about is using it to predict churn. And I think we all agree it would be impactful or it would be important, but we maybe didn't always agree if it would be hugely impactful. I think it was the word impactful that was messing us up there. And then on the marketing content at scale and using synthetic audiences, we didn't actually touch on it that much because it's mid-tier, right? It's not going to be the most impactful or the least impactful, but it's going to be out there. I think with the synthetic customers, I think we're going to have third parties that build synthetic audiences that they license.

Jacob Bourne (21:46):

Yeah, I mean, I had that as my number three for 2029. I think it's going to be very impactful and the technology's getting there, but I think this whole notion of using digital twins for that purpose is it's fairly new. And so I think it's going to get... There's going to need some time for companies to start actually experimenting with it.

Rob Rubin (22:04):

I feel like scientific data, like synthetic medical testing might be really impactful, but helping banks figure out how you might react to a campaign.

Jacob Bourne (22:15):

No, I think you're right. I mean, the whole digital twin simulation use, that started it as a scientific and industrial use case. It's only over the past year that it's even been talked about for marketing. And I think you're right, we do need that synthetic data created first, but I think it will become more impactful for marketing as time goes on, as AI gets better, as people start to trust it for various use cases more.

Lauren Ashcraft (22:40):

Just an example of how much AI has changed just in the last couple of years. In my previous life, I was a marketing copywriter, and we would laugh at AI generated ads because it's like people with 17 fingers and endless loops of cheese when people are eating pizza. And now I saw a video that I would bet all of my money that it was a real cast of actors and a real set. It was 100% AI.

Rob Rubin (23:15):

I know.

Lauren Ashcraft (23:16):

So I have a hard time even imagining what's shortly down the road. Like next year, no idea. But I do know that bank marketers will have to constantly be updating what they think they know.

Jacob Bourne (23:31):

Yeah, no, I mean that's AI right there. I mean, it's accelerating so quickly. I mean, it was just, I think last month that Google released its Veo 3 image generator that floored everybody in terms of the photorealism. But it's a recent release, and so I think we're going to see more ads created using it, and I think it's going to be harder to really distinguish between what's AI generated and what's human generated.

Lauren Ashcraft (23:58):

Totally.

Rob Rubin (23:59):

That's going to be a tough one. I want to thank both of you for today. I had the best time.

Lauren Ashcraft (24:06):

Me too.

Rob Rubin (24:06):

So Lauren, Jacob, thanks for the discussion.

Jacob Bourne (24:08):

Yeah.

Rob Rubin (24:08):

It was really so insightful, and thanks everyone for listening to the Banking and Payments Show, an EMARKETER podcast made possible by Quad. Also, thank you to our studio team that puts these episodes together. All right, next episode is going to be on August 12th, so be sure to check it out. See you then. Thank you everybody.

Lauren Ashcraft (24:27):

Thank you.

Jacob Bourne (24:28):

Thanks.