1029. Insights: How AI is reshaping consumer-focused banking

0:00Today we helped a latte for Sam coffee shop get an insurance quote simply and easily and made sure a floral delivery van was able to make someone's day. We're the Hartford with decades of experience insuring millions of unique small businesses. When it comes to your small business insurance. Thank you. One size absolutely does not fit all. Get a quote or find an agent today at the Hartford dot com slash small business.

0:44Hey folks, welcome to Fintech Insider Insights. I'm David Brea, Group CEO at 11FS. And today we're going to be diving into the big question that we've been thinking about a lot lately. Is AI going to supercharge digital riches? If you're a longtime listener of the show and follow 11FS, you probably heard us talk about digital riches a lot. It's our framework. It's how we define where we really see the direction of travel for the future of financial services and its journey to becoming truly digital. We see this as a evolution of banking to really meet the customer's needs at the point of them needing them and the services that are in real time, intelligent and event driven.

1:24Keeping an eye on what is happening with the client's financial system and ultimately helping clients get the goals and financial life that they really, really want. The framework is an acronym where RICHES stands for real-time, intelligent, contextual, human, extendable, and social. We'll break down these as we go, so no need to take notes, everybody. And now I'm really excited to bring RICHES into 2026 and look at the impact that AI could have on this framework and the opportunity to really supercharge products and services and really fundamentally changed the fabric of financial services. To discuss this today, we brought together an absolute superstar panel thinking to unpack what AI has an impact on the impact for riches as a general and customer experience more broadly.

2:11We're looking at product design, we're looking at risk and resilience and really what it means to be truly digital in 2026 and beyond. First up, we have Ross Gallagher, Head of Consulting at 11FS. How's it going, Ross? first few days back in the office. Sit down and have a nice chat. It can't be too bad, can it? No, it's great. And what a great panel to be a part of and what an interesting topic to sort of kick us off in terms of recordings for the new year. So yeah, happy new year. Well, it's a good one to see if our brains still work or not. Exactly. I think mine's mostly cheese at this point.

2:43So we'll see what happens. But how's your start to the year been? Yeah, busy. Busy as ever. I think it's going to be a really interesting year. I think in no small part due to some of the things that we're going to talk about today as well. So I'm really looking forward to diving into the topic and sort of, yeah, discussing it. Very good. All right. Also, we have joining us Fernanda Doble, who is the product director at Clio. Welcome back to the show, Fernanda. How are you doing? I'm great, thanks. Good. Tell us a little bit more about yourself and tell us a little bit more about your role. Sure.

3:18I'm product director for Chat and AI at Clio. And I've spent the last 10 years or so building behavior change products, D2C. So education, healthcare, and fintech. The last year and a half here at Clio, focusing on building our AI financial assistant. Very, very cool. And it's the first few days back in the office for you as well. How's those trips into the office been? You know, getting back on the bicycle, literally, and into the cold. But it's been okay. Yeah, it's crazy. For everybody listening to this, London, the UK, snow, weird. It was minus eight when I got up this morning. Like, who would have thought it?

3:54But anyway, all right. And next, we have a welcome return to the show, Megan Kaywood, CEO and co-founder of Kaywood.io. How's it going, Megan? Yeah, it's really good. Thanks for having me on, as usual. Love being back. It is always a pleasure to have you on. I mean, you're in the mix of this as well, like you're founding a business based on AI. This is a topic sort of very close to your heart, I imagine. Yeah, absolutely. I mean, my entire career has been in financial technology and I think AI is the next big wave of transformation within, well, you know, every industry, but this space included.

4:27So it's really good to be working on it from the point of a startup because as much as I love working in startups as well as big corporates, it's a bit faster and exciting to, you know, innovate from the startup angle. So yeah, it's a lot of fun. 100%, yeah. Yeah. How's the start to the year gone? Have you managed to get yourself back into a sort of a work routine? No, my husband and I took a full two weeks off. So today is like both of our first days back. So yesterday I had a girl's day with my daughter in London because she just starts back at school today. So we went to afternoon tea and to see the nutcrackers.

4:58So yeah, I have not been working at all, but it's been really pleasant. So that's good. Afternoon tea and the nutcracker suite. I believe you're more British than I am at this stage. I really am because I have my citizenship now. So when everyone asks me, like, where are you from? I'm like, England. They're like, no, really. I'm like, no, I'm a citizen. Well, we're going to have to work on that accent a little bit, Megan, but then. Well, great to have you on. All right, everybody, that is our panel. Listeners settle in and let's get into this one because A, we're all just excited to chat to each other.

5:27So this might run really long, but B, it's going to be absolutely essential listening. All right. So if we maybe dive in and explore a little bit more in terms of where the AI can actually have an impact across all of those things. So like we said, riches for us is about being real time. It's about being intelligent. The intelligence of using all the information, the data that people have, it's using context. It's the context of where somebody is or what time of day it is or anything else that we know about it. It's the human part of it. It is about talking to people like people or talking to people in the way that they want to be engaged with.

6:03It's extendable. It's taking the information, taking the services to where they are. And then fundamentally, it's social. And when we say social, we don't mean look at what I've just bought, posting it onto LinkedIn or whatever. It is about money fundamentally being a social construct in terms of actually how those things work. So I guess maybe starting from the top then, you know, real time, you know, this is a fundamental in a digital age, isn't it? Where, you know, previously where we've looked at, you know, old systems in many organizations that I'm sure all of us worked at, you know overnight batches with the the sort of raison d 'etre and and actually the idea of doing anything in real time is uh is sort of a bit of a flights of fantasy uh and that was not too long ago i don't think i don't think i'm sort of speaking out of turn when i say that's where most organizations are today um but i guess starting with you ross i mean what do you think how does ai have an impact on the ability to work in real time i think potentially huge right and i I think, I don't, I mean, look, there's the danger that I think we all sit here and we sort of wax lyrical about, you know, how AI is going to fundamentally transform everything that we do.

7:12And I think absolutely it has the potential to do that, at least purely from a technology perspective, you know, obviously looking across lots of different sort of silo databases, which we know is a problem in financial services, being able to pull out the salient points, being able to surface those. at the right time, in the right way, I think truly there's a transformational potential there. I think if I step back a little bit, I love the Riches Framework because it sort of makes it quite tangible to almost understand where you're at. Even if I just put AI to one side for a second, sort of in a digital context, moving away from that sort of analog world, I think one of the things that we would need to see in order to actually fully sort of realize the potential not only of AI, but sort of digital as well in terms of building on that is a really sort of progressive approach to how you sort of embrace it, how you sort of roll it out.

8:17and, you know, making sure that what we're talking about is using AI in a way that is additive, that creates a richer experience for customers isn't something like some of how we've seen maybe digital rolled out, which has just been a sort of cost-cutting exercise and maybe a diminished experience for customers. Yeah. Yeah, and I guess, Megan, I mean, roll your brain back to the days at Starling. I mean, you guys and Monzo were some of the first to roll out things like real-time push notifications. And the impact that that has on people's experiences by being engaged in real-time with their financial services is fundamental, isn't it?

8:59No longer is the statement something somebody sort of dusts off and looks at on a Sunday afternoon. But actually, you're being engaged in real-time with your finances as they're happening, right? Yeah, absolutely. I think, you know, if we roll back to like 2016, when Monzo and Starling and everyone were first coming to market, our main focus was exactly that. Like, how easy can you make it? I mean, actually, even when I was in San Francisco and Intuit, we would have this vision of like time to pie in less than five minutes. I don't know how we came up with time to pie as a saying. But everything was like, how fast can you make onboarding while not losing customer trust effectively?

9:34So they know it's like the full experience they need. So it's like you look at onboarding, you look at payments, you look at, you know, downloading statements and, you know, whatever it is that you need to do, just make it as quick and as virtual as possible. That was like the goal. So real time. I think the interesting thing with AI is that real time is still absolutely a fundamental goal of what it can achieve. But it's no longer just about informing. It's about acting. It's the execution of it, the agentic side. So now it's not just how quickly can you open an account? It's like how intelligent can it be about understanding what you need, which account you need to open and access and actually doing it for you.

10:10And so that's the real shift that we get now is it's not just as quick as it can be from a manual perspective. It's how much it can be delegated to the AI to make it a better process end to end. Yeah. And I'd say that probably plays into the work that you've been doing, Fernando, in terms of all the work that you've done with Clio. I mean, a big part of real time is that you're allowing people to engage with, you know, back in the day of like, you know, bank branches being open from, you know, nine till three and never when you need them. And, you know, all call centers open to a certain time.

10:43Actually, you know, chat interfaces and being able to engage with things, doesn't matter if it's 3 a.m. or 3 p.m., then actually the quality of service, the real time experience can be fitted around how the customer wants to engage, right? Exactly. And I think what's really interesting is with AI, you have not just access to what's just happened, but you're predicting patterns on what's about to happen. Right. So it can be something as basic as this recurring bill is about to drop. Let's make sure you have enough in your account to something more complex and behavioral. Maybe you spend too much on deliveries the last three Sundays.

11:16Like, let's talk about whether we want this Sunday to be the same or whether we want to put that money somewhere else. And we find that really exciting potential. Yeah. And I guess connected to that is the second one is intelligence, because actually, the only way you can do that, you know, we always sort of, we've all had these experiences with our banks of, you know, phoning up and, you know, 50 different pieces of information and then realizing the person you're talking to can't deal with the business bit that you've got as well. and potentially they've got to hand you over to another department to do that.

11:47It's like, actually, they don't use all of the information. They don't have that confidence to give you in the way that you described it, Megan, around giving people that balance between efficiency and trust, the confidence that they're there to support you. So, you know, using that intelligence is like a building block of, I think you hit the nail on the head, Megan, trust, isn't it? Yeah, absolutely. I think one of the key things about AI really is that trust and the guardrails that you build in. Because it's not just about making a fast decision and making things faster, but about making the right decisions and the best possible decisions.

12:24And that's where it's always been about trust, no matter what, when it comes to speed and financial services. Like, I remember one of the first onboarding experiences I worked on was for QuickBooks Payroll. But it was really funny. We took out everything you absolutely had to have in order to just make it the most seamless, fast experience. and we had just as much attrition in the conversion funnel as the law experience. And when we ask people why, they're like, it's because I know you need to ask my social security number and you didn't. So even for QuickBooks, like this esteemed, like multi-decade old company, people lose trust.

12:53So similarly with AI, I think it's not just about like, I want speed of execution and you to help me understand what I need to be able to access it in real time. It's I need to have confidence that you're making good decisions and right decisions. And then I can trust you based on the policies and guardrails that I know we have in place. Yeah. Yeah, it's interesting, isn't it? when you sort of look at those first two, Ross, I mean, every big, and Megan, you touch on this a little bit about why potentially it's easier sometimes to do these things from scratch rather than from, you know, within a gigantic incumbent.

13:22I mean, building out real-time systems, having an event-driven architecture, and then you consolidate in and structuring your data in a way that you can use all of that intelligence. I mean, I'll come back in 10 years, like, see if you finished, you know what I mean? Like, it's an interesting dilemma there, isn't it? Because there are two huge, gigantic transformation programs in any bank anywhere around the world, aren't they? Yeah, no, absolutely. I mean, and, you know, fundamentally what we're talking about right now is legacy, right? You know, I think you're talking about legacy technology.

13:53I think you're talking about legacy sort of organizational structures. Really interesting though, Megan, something that maybe I hadn't thought about on quite the same level until you said it just now was the sort of legacy that surrounds customer perception. And when they're thinking about, all right, well, what should this journey look like? And I'm sort of, I have a process roughly mapped out in my head because I've gone through this, this sort of process before. And actually, okay, they didn't ask me for these couple of key points of information. And so I don't know if I really trust this process and I'm going to sort of drop off.

14:26That for me is a really interesting angle here. And so if you go back to the, I suppose, the sort of intelligent, the contextual, well then how do you I suppose sort of frame that journey for customers give them the information that they need that sort of sits alongside the question and what you're asking for what you're not asking for um so that it gives them that confidence and starts to build trust throughout that journey yeah because I think something that's interesting just to add on to that when we talk about legacy when I first went into working at Barclays my mindset was really around the technology and user experience the legacy technology the thing that really needed to shift in my perception was like the legacy ways of working and frameworks and like just, you know, compliance documents that we had for how things had to happen.

15:11You have to shift like both the what and the how concurrently. And that's why I think larger, more established financial institutions could be so difficult to shift because they have to shift so many things. It's not just the tech. It's also the way that they have a relationship with the regulator and what they promise they're going to do and how they're going to do it and how many people agree that we're now going to do it differently and we can capture less and do it and communicate it differently. So it's a really neat problem to solve. 100 % with that, yeah. I always refer to it as the Bananarama principle.

15:38It's not what you do, it's the way that you do it, right? So the way that anybody does anything is usually what results in the outcome. And actually getting that wrong leads to all of the bad things. And I think that's why humans, essentially, every organization, doesn't matter how big they are or small they are. And actually, it doesn't matter how much AI or how little AI. It's like shifting people's behaviors and shifting people's beliefs is very difficult. But Fernanda, do you have any thoughts on that? Yeah, I think it's an interesting one because I think an important distinction we think about a lot is separating the AI as in the products that are using Gen AI or ML systems that we're building.

16:20And then how we're bringing those workflows into our own ways of working. And, you know, in the sense that we should be dogfooding, right? And startups eat your own dog food, make sure you're using the product. But also make sure you're using these systems to understand, because just to pick up on another thing on trust, I really believe we're living through a fundamental paradigm shift in how people change their interactions with technology. We were just talking about voice, right, as being a new paradigm. How do you build trust when you can't look at a screen? And as those shifts happen, how is that going to impact the way companies build trust, the way consumers shift their expectations of what we're building?

16:55Yeah. And that is fundamental to how we are as humans, as you say. It's, you know, how do we interact best? Actually, you know, is it written? Well, written is pretty clinical, isn't it? And actually, is it written in a certain tone for us to break down and understand? I mean, that's a big part of the barrier of financial services right there, right? It's a lot of people using fancy words to make them feel clever, you know, or it's a fear of the regulator to make sure that actually they don't get into issues later down the line. But actually, you know, human consumption of communication becomes a real problem.

17:33And, you know, you've got a challenge there. Do I am I more scared of the regulator or am I more scared that an organization is going to come along and do it well? And therefore, my customers are going to leave to a, you know, a more human centered communication business. I'd say, arguably, that's why all the fintechs have been so successful, really. You know, there's some features and there's some capabilities. The fintech challenge banks, particularly. There's some great features, some great capabilities, but most of it is just talking to humans like humans, really. And back to that point on trust, Megan, building that out again.

18:06Well, something I just thought of as Fernando was speaking, actually, that's quite interesting, is so when I started Starling, I remember Anne would explain that she went through a period post-financial crisis where she's a bit embarrassed to tell people she was a banker. Because, you know, bankers were not perceived very positively. They were kind of seen as, you know, not having people's best interests at heart. And, you know, five years I told everybody I worked in IT. It was weird. Yeah. You know, and I kind of like, I forget that sometimes because I kind of miss that era of banking. But I do think there is still a lingering element where sometimes we mistrust banks or like humans.

18:40But actually technology, I think people generally expect it's not a mistrust of the technology. They expect it to be objective, but it's more around just ensuring that it understands you, it has guardrails in place, that you can see what it's doing. At any given time, especially with something that's automated, you can see why it happened, when it happened, you could reverse it if you need to, that sort of a thing. But I think internally, one way that AI within this framework will be successful is whenever we take the regulatory frameworks that we've applied to humans post-financial crisis for accountability, like the SMF functions.

19:11You can be a senior manager, you can delegate tasks, but not responsibility. When you assign that to an agent, it's the same thing, but it should have a human steward. You know what I mean? So that there's still an accountability. So if the AI system fails, it's not all blend goes on the AI system and you roll it back. You look at the systems of accountability and how you need to evolve it to continue maintaining, you know, the kind of level of fidelity and standards we have in financial services. That's like trust has to shift now that we have a new way of executing with different agents who are acting.

19:38Yeah, we see that all the time. People are they report in user interviews, they're more willing to trust an AI because they don't feel judged. So it's, there's like the mistrust on legacy banks, but there's also this judgment of like, oh, this successful financial advisor who's got their shit together. I feel like nobody talking to them and actually an AI is not going to judge me. So that feels safe. But I think there are moments though, you know, I think you're always going to need sort of like human touch points and human interventions when it comes to financial services and when it comes to people's money, right?

20:14I mean, there are scenarios that are really scary, like if you're the victim of fraud or things are happening sort of in real time. And this is really good. I think real time example is, you know, how can AI sort of like help to intervene, catch those things when they're happening in real time? But I think when people then are feeling sort of vulnerable, they're probably feeling quite scared. They're feeling quite anxious and worried. you know that obviously then is a really high impact touch point I think where you've got that sort of human intervention I think you're probably also going to need that um going forward as well but it's how do you couple those two things right and how do you make sure that those really high impact touch points are sort of handled in the right way I think that takes us really nicely onto the next point around contextual because I think you're completely right I think the you know the the argument we always make is um you know the behavior the context of what somebody trying to do denotes the way in which to do it.

21:05So, um, you know, sadly, when my mom passed away, uh, loads of banks pushed me towards filling out loads of forms in order to, you know, access it. And it's like, it didn't feel justified. It felt like they were making their admin my problem in a point where actually all I needed was somebody to help me and deal with that. Um, similarly, we will say it's like, you know, the idea of, uh, a loan being a loan, actually somebody applying for a loan at 3am is fundamentally different than doing a 3pm, right? So, so context really matters and understanding the context of the customer, the behaviors that they're exhibiting, the, um, the stresses that might have been coming from, from that context.

21:46These things are really critical to make sure that actually you can, you can serve people in a consistent way or, or serving them in a way that's appropriate for the, for the needs that they really have um what do you think megan yeah i think the interesting thing about this is that you're exactly right context absolutely matters it matters in those key moments like you have you know a relative pass away and you're working through the finances that even as fernanda said like you need financial advice and you're a bit embarrassed i think something i was really fascinated by when i was doing research for kwood last summer when i was first starting out as i was um exploring how people saw you know financial advice what drove them to it and how that process went.

22:25And what I found really interesting is even at the highest levels of earning, a lot of people would avoid getting financial advice because they were just embarrassed to open the kimono. But they would report, like, they don't mind talking with ChatGPT about it because they don't expect to be judged. You know what I mean? And so I think there's this element you need the appropriate advice and the appropriate response at the right time contextually. But more than that, I think the interesting thing about AI is it's getting really good attunement, recursion, emotional intelligence, to be able to understand where you are emotionally, to meet you there in a way that I think we didn't really expect the ability to have a conversation and to be able to respond effectively.

23:04But more than that, I think there's a really interesting dynamic that's playing out where I think it's exactly as Ross said, where there's a relationship between AI and whenever it hands over to a human. And China's done this really well in the US. They have been working on their AI integration over the past year, and they've done it in ops and customer support. And they have a really neat system. They ended up reducing their ops costs by 60 % while also doubling their NPS score. But it's because they have that relationship really nailed down of like, where do we start with AI? And when do we need a handoff to a human where that's the right path to go instead?

23:36And I think that's the future of it because we need context. You need emotional intelligence. You need to meet the person where they are. But then also, if the automated or the agentic experience isn't right, When does a human need to step in to pick the baton up? Yeah, you know, one of the things we use at Clio is this framework around proactive and reactive jobs to be done or conversations. So when a customer comes to us with a very specific job to be done, we need to identify that the AI needs to identify and then use intelligence to send them down the right path. Versus they come in, they're engaged, they've satisfied their job to be done.

24:08How do we then be proactive about showing them that just-in-time behavior, right? And so we think of these as two different problems to be solved and making sure you have the right one at the right time so we can measure their success differently. Yeah, that's interesting, isn't it? Solving a problem that people are aware of versus solving a problem that they might not be aware of is fundamentally different in the way in which we approach the communication around it, isn't it? So, and most of, I guess, in this world of trying to, you know, financial services, I was joking, financial services is a thing that we all care about, but nobody else does.