Redefining Data ROI in Banking: Luke Pearce, Chief Data and AI Officer, Santander UK

0:00The experience I had probably pre-banking has probably shown me that there are things out there in the wider world. I think banking, when you're in it, it feels like the most important thing. There are lots of other things in people's lives. And I think the bit that data can do is actually start to understand a little bit more through, as I say, through that journey of customers are taking, what they're paying, spending their money on, what type of financial instruments they're using at the time. you can really start to tell a lot more about a customer's work or their life outside of those transactions.

0:32And that's where I think the power of the data really comes to the fore.

0:39Welcome to Data and AI Mastery, the podcast where we bring you cutting-edge insights, practical advice, and inspiring stories from the leaders shaping the future of data and AI across the globe. I'm your host, Raul Gabriel-Urma, founder of Cambridge Spark, the leader in transformational data and AI upskilling, career development and progression. In each episode, I will be diving into real world case studies of companies harnessing the power of AI to drive innovation, reduce costs and create new business opportunities. So whether you are an aspiring data scientist, AI engineer or seasoned executive, this show is designed to give you the tools and knowledge to stay ahead in a world where data is transforming every aspect of business.

1:24Stay ahead, stay inspired, stay masterful. Welcome to Data and AI Mastery.

1:33Hi, Luke. How are you? Good to see you. Good to see you. Nice to be here. Yeah, I'm really excited for our conversation today.

1:46Luke, you have a very prestigious position. You're the Chief Data Officer at Santander, UK. Wow. So, I mean, to kick us off, I'd love to get a bit of your history, right? Like when you look back at your career, what were some of the key moments that got to where you are today? Yeah, I mean, I didn't think probably it was a bit more of a squiggly career. as probably a lot of people in this position go through. So I didn't start off even thinking I was going to go into a technical career. So I studied Japanese and economics at university, if you can believe that, and then actually went on a graduate training with Capgemini.

2:28So looking at more technology type project work. And I spent a lot of time building front end systems before I actually even got close to the data side. And probably during that period, there wasn't so much happening in the data world. And that started to build up. And I mean, in fact, I think I probably had some early exposure to the data side because the infrastructure was pretty poor and things took a long time. Lots of things were running in batch and taking 24 hours to run. But as the technology started to improve, I think that's where data really started to come into its own. I think probably some of the key parts for me was I went from being in the non-financial world.

3:09So I did a lot of work in manufacturing. I also worked in corrective services in Australia for a few years and then switched into a financial role doing financial reporting and actually starting to understand the data that banks have. So that was a big step for me. And I think that's where I really started to see the value of how data can be used to drive a lot of those customer and bank outcomes. So I spent four years with Commonwealth Bank and then I came back to the UK around 2006 and joined Barclays. And I was at Barclays for 15 years doing various roles, but more and more on the data side.

3:46And starting to see, again, how you can make data really starting to leverage some of the value of what the data has to offer. And I think, interestingly, where Barclays were at the time, they were starting to see the benefits of really understanding what the customer behavior was, what customer if you think about from a banking perspective you know lots of payments and transactions flowing through the system and trying to see the insights and how you can use that to your advantage so we did a lot of work understanding industry insights so looking at customer spend on the retail side to drive insights on the corporate side and also starting to see where we could really start to engage with customers in it with a different conversation to maybe the traditional banks had so so I think a few things and then and obviously in the last few years coming into Santander again going going starting off with a technical role and then finally picking up the chief data officer role I think I've been there three years in that role now and obviously in that time then starting to look at data and AI which I think has become more and more important in in the data side and making sure that's done with with a lot of control at the oversight as you can imagine yeah okay well we're gonna have to talk about the controls and governance in a moment but it's such a fascinating background right uh you clearly you normally you've traveled the world but uh you've worked across different industries so i do wonder like you know have expectations around what you can get out of data change over the years or across industries and if they did like you know what do you think is the edge that you've acquired you know for all those different industries and experience because it's pretty unique place to be I think expectations have changed massively.

5:26And like I said before, I think the fact that the infrastructure and the tooling has become so much better now. I mean, in the past, we used to talk, we used to dream about having things like customer telephone calls converted to text. And that's become part of the everyday now. So things where probably there were a bit of a challenge in the past have now become part of the everyday. I think what you've also seen is from a customer perspective, the expectation is massive now. And if you think about when we're all online at various times of day, there used to be this annoyance of, you know, I've searched for something on one site and then it follows me around.

6:05Now, I think having the right conversation at the right time is an expectation that every customer has. And starting to use the data to understand that position that a customer is at a particular point in time, I think is really important. I think the experience I had probably pre-banking has probably shown me that there are things out there in the wider world. I think banking, when you're in it, it feels like the most important thing. There are lots of other things in people's lives. And I think the bit that data can do is actually start to understand a little bit more through, as I say, through that journey of customers are taking, what they're spending their money on, what type of financial instruments they're using at the time.

6:45you can really start to tell a lot more about a customer's work or their life outside of those transactions. And that's where I think the power of the data really comes to the fore. So if you're having the right conversation with the customer, you understand what their need is at that point in time, and you're there to support. I think that's where data really is showing its value. That's super interesting that, you know, of course, technology has evolved. And then as a result, consumer behaviors, consumer expectations have changed and are rising up. And as a result, you know, you have to adapt along the way.

7:24So I'd love to continue the thread around expectations and maybe at the both at the executive level and at the board level, because certain organizations think differently about the value of data and what they can get out of it. So I'd love to get your perspective. you know if we start at the you know at the c-suite you know what are some of the organizational structures you've seen uh in order to you know support certain expectations and you know and in your most recent experience like what's the setup it's a really interesting point because i do think that you can be stifled a lot in the data space by people thinking that data is there you have to get under control and particularly in the finance industry there's there's a lot of regulatory screw to the end, you know, have you got your data?

8:10Is it trusted? Do you have controls on it? And you end up in a very governance-focused environment. And I think that's probably the bit that, like I say, stifles creativity and it really ensures that you're not getting the best value out of the data. And probably I would say lots of the banks have been through that journey over the last few years because we've all had risk models. We've all had, I mean, BCBS 239, You know, we've got SS123 now. And so the regulation is coming in saying you need to have your data and you need to have control of your data. What's been really nice in Santander is we've actually then started to see right from the CEO all the way through the Exco is a real, you know, almost an invite to can you make the data do more?

8:59Can you show us the value and how can we support you in that journey? And I think that's a really important for any organization. If you have that sponsorship at the C-suite level, it becomes a lot easier to open the door to lots of those new opportunities. And I think the bit for me, when we started to look at generative AI, because obviously generative AI became one of the buzzwords, everyone wanted to understand it, what it could be used for. But having that support means that you start going from probably an environment where people are very risk averse to doing anything new, because they want to know that the, you know, the regulation is going to be supportive, that we're going to have like controls.

9:34And can we do these things, which, you know, are really quite new and innovative for the organization, you know, without having that, that worry. to actually let's start to push ahead with this agenda. Let's start to see how we can make it work, both for our colleagues and our customers. And it opens up a lot of doors. And in a different way, you have people in second line and third line actually trying to find ways through and trying to manage these new environments where you need to probably identify new controls. You need to do new ways of looking at things, how you manage your risk portfolio.

10:08It's all of those things come together. and actually the outcome is quite amazing. So some of the things that we've managed to do, I think in the last couple of years are probably as a result of that support that the C-suite has given us. That's fantastic. So if I play this back, it sounds like Gerantive AR has almost like provided a bit of momentum and fuel to kind of like get that sponsorship that is really crucial to, you know, make things happen. I guess, what are the expectations still and how do you manage those expectations? I feel it's great to get the sponsorship, but I imagine that you still want to be a bit cautious around, And what are the possible ROI here that we can get out of various generative AI or data initiatives?

10:45I mean, ROI, again, it's an emotive subject when it comes to the generative AI agenda, because I think a lot of people went into that, okay, with generative AI is here, it can replace, maybe you can replace roles, maybe it can improve productivity. So let's look at it purely from a cost driver and how much cost out can we manage with these new models? Again, I think the approach we've taken, we've not really focused so much on just cost takeout. I think there's definitely an element of how do we improve outcomes for customers, for colleagues. I still I mean my whole belief is that there's a lot of things that we do in large organizations which I'm sure it's the same across it's not only in the finance industry but across all industries where people are doing low value admin tasks and that is stifling their innovation is stifling their productivity that's where I think AI can play a really good role so it's about you know if I think about trying to consume large volumes of unstructured data AI can give you that summary We can work on things which can start to manage a lot of the decision making, whether that's when we're starting to build contracting or look at contracting with third parties, whether it's looking at some of the financial crime, unstructured analysis and assessment.

12:04Really presenting that data to someone who is an expert in the area, I think, accelerates that. On the customer side, I also think there are things where you can take a customer through a journey that maybe took three weeks before and you're now doing that consistently in like less than a day. Those are things which really start to put the value into this. As a result, there might be some cost takeout. And like I'm not saying that that doesn't come part of it. But if you don't have a balanced scorecard across that, and I think this is where the expectation across all of these things is, I expect to see improvements for customers like NPS.

12:38Yes, I expect to see more productivity for colleagues and colleagues actually embracing the use of AI. And as a third point, yes, there is cost takeout. But all of those things balanced together means that you pick the right use cases. And then you start to go on things which actually start to really improve the organization as a whole. And you free up capacity for probably those areas where you probably in the past haven't treated them with as much time or effort as you needed to. because you've now managed to do that a lot of the things which were probably out of reach out of scope before and now coming into scope and overall as an organization really starting to improve across the board i think ai has been like fundamentally yes it starts with a with an expectation driver um but we're meeting the expectations along the way in multiple disciplines which i think is great that's really powerful what you're saying and i love the the focus on I guess customer delight, employee delight as a North Star.

13:35And by focusing on that, you're absolutely right. There's things like time to response. There is fraud prevention and so on, which have side effect and cost reduction. But that's not the primary driver, actually focusing on the delight and happier customers and happy employees, actually, longer term, that pays off. So that's really interesting to hear that. Now, you did mention earlier the risk and control, right? So I guess the question I have for you is, what differences are you hearing in expectations between the C-suite, the X-Co versus the board? Do they think differently? Are there different concerns from your experience?

14:13It is interesting because I think even at the board level, there's an appetite to use AI and to start seeing how those things can be used. Now, obviously, that in itself comes with a lot of challenge. But I think it means the conversation is easier because this is something, and I've talked about this a lot with people on AI, I believe, will become like digital. So, you know, when we started putting things into the app space, people were on their mobile phones. It was a thing in itself. And we all set up kind of digital capability. And then it just became part of the everyday. I think AI would be exactly the same.

14:49and because people are engaged with it outside of the working environment they come in you can have a conversation with people and they they start to understand it but they also start to understand things like hallucination and you know what can go wrong so I think the conversation with the board on the one hand which is yes we want to get more involved we want to start to use it on the other hand it's how are you showing that control and that's that's a really difficult thing to answer at the moment you know without having to understand you know how models work and what you're doing, you've got to break it down into something which is, which I think is consumable by people working at that level.

15:25And what we've tried to do is, is, is look at, because I, like for me, I think the idea that you, you keep talking about, you know, the, the hallucination and the one-offs and the outliers is, is the wrong conversation to have. And I've, I've said many times that if you have, people will go onto our, you know, chatbots on various things. I was on a chatbot on the weekend trying to get, um, you know, some stuff, sorry, with a with a like a mobile provider the chatbot was awful and it gives the wrong answer but no one really gets you know because eventually you get put through to a person no one really minds on it as soon as you get into an ai chatbot which gives a funny answer then everyone says oh it's hallucinating it's going to give like all sorts of bad things and and so there is a panic almost setting in it's a very emotive subject and i think we need to move away from that guardrail discussion on, you know, we will manage that it will produce a good outcome, as in it won't produce something that's hallucinogenic.

16:21But we also need to show you the controls that we have in place across the flow to make sure you understand that and you understand that actually we're monitoring these things on an ongoing basis. And the way I try and break it down is, you know, operationally, we need to make sure that we're calling these things in the right way. So you can look at, you know, if you're linking AI together and you're calling a model in a number of different ways, if it starts to do it, the patterns start to evolve and start to do things differently, then that's worth an investigation. So we call that the operational metric.

16:51We also need to make sure it's being fed with good data. And I think that is one of the most fundamental parts of the crossover between the governance side and the AI side is if you put bad data in, it's almost exponentially worse than what it used to be of what comes out the other side. So make sure you have your data controls. And then the third part, which is, let's look at outcomes and let's make sure there's an oversight in terms of making sure that we, the data we're feeding in is producing the outcome. And we're looking at the outcomes and checking it. And that could be, I mean, it could be some partial manual checks that people are doing.

17:23There could be AI as a judge and those things will start to evolve over time. But I think those three key parts of making sure that you can describe the way the oversight is working is a much better conversation to have with a member of the C-suite, a member of the board, so that once they understand that, then they can see, okay, well, how many things have we got live? You know, what are those, how are those controls working? And do we have confidence that they're working across the organization? I think that's the only way that you scale. And that's the only way that you then start to get into a conversation around agentic or something a little bit further, which is, okay, can we put this now in front of a customer?

17:57Do we trust it enough? But it is really that core conversation I think is so important at the moment. I hope you're enjoying today's conversation. If you're finding the insights useful, please do take a moment to subscribe to the Data and AI Mastery podcast and leave us a review on Apple Podcasts, Spotify, or YouTube. Every new follow helps us reach more people and shed incredible work being done by today's Data and AI Leader. All right, let's go back to the episode. Super cool. I mean, hearing you out, you know, and I speak to like lots of organizations, including financial services, you're quite ahead, right?

18:33You've got a board that is on board. You've got a C-suite that is, you know, like really committed. You have a clear strategy focusing on customer delight in terms of value creation. you've got your validation framework processes and regulatory sort of a framework around that many organizations are not there yet so i do wonder what some of the the key lessons of what's working what's not working that you know you you could provide us guidance to organizations that you know have the ambitions to get there but you know are just getting started i mean we did take like quite a good approach at the start so i think a lot of a lot of organizations you know they they the AI is there, they say, let's find some use cases.

19:14And you get a list of a thousand use cases, and it's very difficult to digest them all. I mean, I think probably everyone goes through that first process, and then you're having lots of conversations. You're not actually progressing anything because you're trying to work out what that killer use case is. I think we took an approach early on to bring a domain approach together. So we effectively, rather than say, individual, anyone can bring a use case. We said, let's focus on a domain. Now, as we all know, when you build a model, that's only part of the story. So if you're going to build a model, you have to integrate it into a front-end system.

19:51You have to then bring it into an operation. Those are fundamental parts and they can often take longer. And equally, you've got the governance running alongside that. So the point of that approach was to say, we probably have a limited number of core systems in each of the domains. We have an organization where if we start to build an understanding of how the AI will work and how it will affect them, it becomes part of that cycle that they can get used to doing training, understanding the models, understanding their role to play in the process, and really building teams that cut across the different parts of the organization.

20:32So we're talking about pivot to agile at the same time, but it really doubled down in terms of the AI flow. So we had the business, the architecture teams, the technology teams, the data teams, including the data science and the engineering teams, all working together and effectively building out a roadmap for how we would land AI. And so then it becomes much less of a surprise because you have teams that are ultimately, and this is where I think AI is really interesting, teams that understand the regulation, if it's like a financial crime, people who understand the operation because they're working on it day to day.

21:11They are key parts to shaping the way the model works and bringing them closer to the data science team. And the technology teams means when it rolls out, then it rolls out in a much better way. So I think that was a key part for us. It's really about making sure this isn't something that is driven by either technology or data and us assuming that the answer is going to be perfect. It's about making sure we have that joined up team approach and really being challenged probably by the business and the operations teams from start to finish to make sure that what finally gets implemented is going to be successful.

21:50and equally I would say that even when you think you've got the perfect solution you know doing that first round of testing and making sure you know it's it's probably rolled out with a a few really good experts in the organization means that then you really start to do the the live testing you know with oversight with challenge and so when you get to the maybe with a scale of rollout, you really are in a good situation. And it's only with that skilled individuals who are business SMEs that that really works effectively. And so I think that was probably the best part of our success. Amazing. So I'm hearing the importance of getting the business engaged, the operations engaged, collaboration across teams, have a shared success criteria.

22:41I can't help you to think, when it started, you know and you're with the ceo and you have to pick a domain or you have to pick you know this kind of like instead of a thousand projects and pilots you've kind of like to pick one that we're going to put a weight behind like how did that happen like how do you pick that domain and how do you balance effort versus reward because i imagine you know you kind of need to show results quite rapidly to keep the momentum i mean i think we we probably picked a like i won't lie to you Like I think sometimes this is you as a personal, you know, it was probably areas of the business where people really could like start to see the value.

23:21I think there's always you always want to do something which because we didn't just do one domain. So we did. I think in the end we did about four. But you always want something that's a little bit more customer facing. So customer interactions, which is our contact center, our branches where we'd already rolled out a knowledge base. because I think a lot of the organizations, that's one of the first things that you do is bring the knowledge base to the colleagues so they can start to surface things much more quickly. And we'd had some success with that. I think we'd done some areas which were more operational in nature, again, from a customer perspective where you're starting to manage lots of unstructured data.

24:01I think one of the greatest things from an LLM perspective is turning that unstructured data and large volumes of information into something that's structured and really consumable. Those were two areas which were, I think, quite easy to decide upon. And then we had, I mean, equally from a customer perspective, the financial crime and fraud side, which probably had a lot of data expertise already involved. And so it was a much smaller step to make them, you know, kind of look at the AI and see how the AI was going to add value. but yeah there's always um you know and and one thing i would i would say as well is you go down this path it's it's not perfect i mean i can talk i can talk a good game but when you start to build your road map so this is one thing that's that's really interesting is there'll be things which will um fail miserably and very quickly and you realize actually what you thought was going to be a great use case gets destroyed like in about the first week but then what you also find is some use case that you thought it was quite easy to do and showed a little bit of value can then be reused across more and more parts of the organization.

25:10So that knowledge base, which started off just being about information from the bank, which I would say colleagues and would use in the contact centers in the branches, has now become a knowledge base that contains some of the policy information and it contains things that other parts of the organization are using because we fed more and more data into it. And now that's probably one of our biggest successes across the organization and probably has more users than any other AI use case. So it is that bit of it's bumpy. And that's quite a difficult conversation sometimes to have with people who are investing in you to say, well, actually these five haven't gone so well, but these other five have done much better than we expected.

25:50And I think that's the way the AI journey goes in a lot of respects is, And it is trying to double down on the ones that are more successful and trying to stop the ones that aren't going to be successful as quickly as possible. That's fascinating. It sounds like you do need to keep an open mind, be experimental, think about innovations. Some is going to work, some is not going to work. And this is kind of like a learning process almost.

26:21can i take you to a quick fire round of questions like short questions short answers sure super all right maybe the first one is what's a contrarian view you have in the industry is this something that you know you disagree with you know that you want to challenge i feel that the job of the CDO will become much smaller in the future. I feel like AI will be everywhere. So my challenge to people is, if I was thinking about hiring someone to run part of my business in the future, would I hire someone who is AI first? Yes, probably I would. Would I hire someone who was a data scientist? That's maybe not quite there yet, but that's my honest belief is, and I've kind of painted this as the death of the CDO is my job becomes much more redundant when people know where they're taking their data from because the AI can manage all the oversight of the data it becomes much more redundant when everyone knows about AI and they're managing and they're building AI in their own parts of the organization cool well thank you well that's a very cool view thank you look the next one is if you had a magic wand and you could fix you know anything at the organizational level what would that be so i think at the moment i sort of feel um again if i talk about ai the conversation is you've got people who want to get involved in it and and can't get involved in it so i would make it accessible to everyone um i know that's not always easy The other bit that I would try and remove, if I had a magic wand, is fear of AI.