AI Delivers Hyper-Personalization and Governance at Esquire Bank

Dylan Cosper:

I'm Dylan Cosper with the Emphasis Knowledge Institute, and today I am joined by Kyall Mai from Esquire Bank. Kyall, thank you for joining us.

Kyall Mai:

Thank you very much for having me.

Dylan Cosper:

So could you give me a little bit of a background around Esquire Bank, some of the things that y'all are doing right now with AI and marketing?

Kyall Mai:

I'm the chief innovation officer at Esquire Bank, and Esquire Bank is publicly listed on the Nasdaq, and we're a digital only B2B bank, specifically targeting the litigation vertical, and the payment processing vertical as well.

And some of the things that we're doing that's quite interesting at the bank is we're really heavy on B2B content marketing, and using AI and hyper personalization to really send people the right message at the right time. And we're doing that at scale, at a national level.

Dylan Cosper:

I'm guessing a lot of those efforts are centered around maybe some localization of ad content, things like that.

Kyall Mai:

100%. It's all about hyper-personalizing content, and using generative AI to supercharge the content creation capabilities.

Dylan Cosper:

We were, just before this interview, discussing the event that previously happened with ANA and New York. And one of the things that came up was intelligent brands. You now no longer have just one Esquire Bank. You have maybe hundreds of different Esquire Banks depending on who your target audience is. What excites you most about that future of branding and marketing with AI?

Kyall Mai:

Yeah. So at its core, what we try to do at the bank, and this is what makes life difficult, but a really good challenges are always, our target audience are 55-year-old lawyers. And we're trying to convince them to switch banks from a relationship they've probably had for decades, and sometimes even multi-generational.

So using AI to connect with a person through smart content, content that they love, to build brand affinity. And to somehow, because we don't advertise on television, reach hundreds of thousands of people digitally and build that brand affinity is really what excites us about AI, because the capabilities are very different today than what it was before from a content creation perspective and from a targeting and segmentation perspective.

Dylan Cosper:

Now, with your current stage and where Esquire Bank's marketing is with AI, is there anything that kind of surprises you about the stage you're currently at?

Kyall Mai:

Yeah, I guess the first thing is, given we're a highly regulated industry, I'm surprised I'm allowed to do it. We went through quite a lot of effort to ensure that the bank understood the safety and soundness around what we were building around our AI governance and our platform.

And we put the right things in place to ensure that our customer data was safe, and all of the technology and security standards of the bank were safe. So from that foundation, that then allowed us to do quite a few things in the marketing space. In fact, we're the only department that's been allowed to use AI for now, but we're using it as a test case for other departments, because they're looking at what marketing are doing and they're finding some of the things that we're able to do quite interesting.

Dylan Cosper:

And we have some previous research. There's the CMO Radar 2024 that found that, for many companies, the CMO and the marketing function is really that test bed for new use cases where you can find value with enterprise AI. So it's really interesting.

Kyall Mai:

Hundreds. Especially with... Let's start with generative AI. Because marketers deal with words and images day in, day out, synthesization of words, creation of words. Now, if you think about other departments like the legal department, the compliance department, they all use words.

So the things that we learn in marketing to master the utilization of words with artificial intelligence is globally applicable to the rest of the enterprise. And marketing are this nice safe test bed, because there's always a human in the loop. Before any words go out to customers or ads, multiple people are checking it. So that's why our organization felt that it was a really good place to start.

Dylan Cosper:

And you mentioned having a lot of these controls and processes, and I guess safeguards in place that have enabled y'all to really kind of take off. Could you talk about your approach to responsible AI at Esquire Bank, at least as it pertains to marketing?

Kyall Mai:

I think understanding that we're at the very early stages of this potentially groundbreaking technology, and there's a lot of things that people are unsure of, starting from that base of getting that there's still a lot more to go and there's still a lot of changes going on. There's still a lot of actors in place.

When you start from that position, the first thing that you do is you think, okay, you need a governance model in place, in order to properly assess what the opportunities are, what the risks are. So we have an AI governance committee at work that's comprised of myself and compliance and legal and InfoSec representatives as well.

And they're really there, not necessarily to just stop everything, but to just ensure that people are doing things in a safe and responsible way. I'd like to say that we are cautiously optimistic about AI, and I think that's a really good way to approach it, because if you go too far running ahead, you might get caught with your pants down eventually.

Dylan Cosper:

Brand trust is incredibly important. So would you say that's maybe the biggest thing for y'all when it comes to responsible AI, is that brand trust that you don't want to lose?

Kyall Mai:

There's a responsibility to the organization, yes, from a brand trust and reputational trust perspective. But from a human perspective of the people that work for your organization to understand how this technology could potentially improve their lives or impact their lives or change their jobs, that's really one of the core tenets of the AI governance committee as well. It's not just, let's figure out our technology risk framework. Let's figure out how this is going to impact people's lives for the better, in hopefully in most cases as well.

And let's go through and have some of the questions around training and business change. And that's critically important and a big role of the AI governance committee.

Dylan Cosper:

Currently, where are you and your team seeing the most value from AI? You mentioned content, a bit about personalization, but are there any other areas and maybe some specific use cases that you've seen most valuable, at least today?

Kyall Mai:

So, we have gone all in on personalization and AI to help decide on what content should be served to what person. And that's because there are many different types of lawyers that we target. For example, there are divorce lawyers, there are personal injury lawyers, medical malpractice lawyers, and they're not interested in each other's content.

So, in order for us to convince some of these lawyers to switch banks, we need to convince them that we understand their business and we understand what they do. So we used AI for hyper-targeting messages and content. And then on the other side of the fence, we also used AI from a generative AI perspective, to help us what we call, "atomize content."

So as an example, we don't think AI is quite there to build fully production-ready content, especially towards lawyers. You have a high expectation when it comes to what they read, but it doesn't mean you can't write a piece of content and have AI atomize it into 50 different pieces of content underneath that, 15 LinkedIn posts, emails and all those kind of things.

So starting from a core piece of human written, trusted content, and atomizing it into many feeds our personalization engine very, very well.

Dylan Cosper:

Now, how has this value and these use cases compared to what you expected to be possible to date, maybe last year or the year before, with AI? I mean, is it meeting expectations or is it far exceeded?

Kyall Mai:

It's definitely meeting expectations. And it's also, at least from a content perspective, it's interesting how often models are changing, even just yesterday or the day before Google Video came out and you're able to generate videos. We haven't quite gotten down that space yet, but I could imagine that's going to be a game changer when, in a similar way to how you can create content with words, you can do that with video. That's going to be a very interesting world for marketers, that's for sure. And we're looking forward to that.

Dylan Cosper:

Where has AI, at least today, kind of fallen short? Where would you say?

Kyall Mai:

I think in terms of being able to get the same results over and over and over again, so when you work for a bank, sometimes you want consistency over creativity in certain scenarios, which is where I can imagine some of the other departments might struggle a little bit. But with marketing, it's a bit more moldable like Play-Doh, and you'll take creativity over consistency.

Dylan Cosper:

For sure.

Kyall Mai:

So looking forward to changes in that. And I guess at the enterprise level, there's AI when it comes to AI agents, and then there's AI when it comes to individual use cases and people using generative AI. Don't think my expectations have been met from an agentic perspective yet, but we're still in the very early days. And even ourselves, we're kind of just playing with it right now, but I have a feeling things are going to change in the next couple of years very significantly in that space.

Dylan Cosper:

One of the things that we're finding in other conversations has been how important experimentation is. Because at the end of the day, there's maybe some unknowns about what AI can do or how we need to govern it in some regard. But how important is that experimentation to you and your team, and what kind of frameworks do you have set up to enable that experimentation?

Kyall Mai:

Curiosity is probably one of the best things you can have when it comes to utilization of at least generative AI, and especially in the teams itself. And fostering the courage to try and do new things and not trying to funnel people down some sort of predisposed template.

Because it's tempting to have, okay, we have a prompt library now, let's everyone use the same prompts and then everything becomes the same. But fostering courage and creativity to do different things is critical.

And also rewarding, not necessarily what happens when you save the time by using generative AI, because that's going to happen. You're going to shave five minutes off, 10 minutes off, 15 minutes off here and there, and then all of a sudden your staff members are going to have an hour or two to spare. Don't reward that. Reward what they do in that space, the creativity they do, the collaboration they have with salespeople, that phone call that they wouldn't have made before because they are too busy, but now they're making that phone call and building that relationship.

So, somehow validating those things that they do is important and fostering more of that to happen.

Dylan Cosper:

I would like to go into more of your current challenges.

Kyall Mai:

Sure.

Dylan Cosper:

I mean, I think everyone has them. Some of them are the same, some of them are uniquely different. You've mentioned your unique situation with lawyers. What have been some of your challenges and your team's challenges with adopting AI within the marketing organization?

Kyall Mai:

It's mostly the pace. Things are changing so quickly, and it's very tempting to jump onto the next shiny new toy, but equally at a regulated company, you can't necessarily use the things you want to at the pace that you'd want to. So it's this interplay between, "Oh, I really want to do these things, but it needs to go through compliance and all that kind of stuff."

So it kind of slows you down. But equally running forward towards some of those things, one of the challenges is because things change so frequently and the training aspect of it is big as well because the big question is how do you share the skills and knowledge of some people that are excellent at AI across a large group of people and do that at scale? And that's definitely one of the challenges that you have at probably most enterprise organizations.

Dylan Cosper:

And so on that training and upskilling talent, I mean, what have been some of your approaches to addressing that challenge? And have you seen improvements, anything that's been more successful? You mentioned the experimentation itself, but are there more formalized processes for the upskilling in AI?

Kyall Mai:

Yeah. So, we currently have small pilot teams using AI, and we have a monthly professional development and training session. It's formal-ish, where a lot of the times I'm talking about what I'm doing with AI and the interesting use cases I'm using it, just putting a lot of effort into the actual training face-to-face of people as well, and they love that.

Dylan Cosper:

How would you describe Esquire Bank's data challenges with AI adoption and marketing?

Kyall Mai:

Data is absolutely the cornerstone of success for AI. Luckily, we invested in data many years ago. Not necessarily because of AI, but mostly because we're a bank and we wanted to achieve single customer view and have good quality data.

And from that platform of having great data in our CRM system, customer data as well, it gave us the confidence to do things, for example, with Salesforce, to implement the data cloud and look at their AI solutions.

And what that means is we don't have to look backwards and try to fix things. And instead, we have this beautiful foundation of great data that's been unified in the data cloud in Salesforce, which can then bring forth innovation cycles across many different departments. Innovation cycles that we don't even know of yet, but we're confident that from those foundations of great data, we're going to do some amazing things.

Dylan Cosper:

Kyall, I'd love to turn it over to you for any final thoughts. If there's anything that we weren't able to cover in our cnversation so far that you would like to share, by all means, the floor is yours.

Kyall Mai:

We really are at the beginning. When I think about ... So I'm 46, so I was lucky enough to be part of the internet when it first came out, part of social media and all those things.

And if I look back 25 years and see how much things have changed, it's going to be the same with AI. And AI is going to be even worse, because you've got AI teaching AI and improving on AI. So, it's going to be a brave new world and I don't think it's all doom and gloom, but we all going to figure it out and meander our way through there, and hopefully with some positivity and optimism and guardrails, we'll get to a better place.

Dylan Cosper:

Absolutely. Well, thank you very much, Kyall.

Kyall Mai:

Thank you very much.