How Nationwide taps Kafka, MongoDB to guide financial decisions

Emerging technologies and a fast-changing competitive landscape mean financial services firms must explore new business models, according to the analysts at research firm Gartner. That’s certainly the case at the U.K.-based Nationwide Building Society.

CIO Gary Delooze is using data-led digital transformation to help the organization explore new ways to serve its customers and support long-term business growth. By taking advantage of data technologies like Kafka and MongoDB, Delooze wants to give Nationwide customers deeper insights into their finances.

Delooze says modern businesses must have a strong data model in which employees understand what data is stored where and how this information flows through the organization. He explained to VentureBeat what this effective data model is allowing his team at Nationwide to achieve on behalf of its customers.

This interview has been edited for brevity and clarity.

VentureBeat: What do you think an effective data model looks like?

Gary Delooze: A lot of organizations try and go for a big data approach — let’s throw everything into a data lake and try and capture everything and then work out what we’re going to do with it. It’s interesting, but actually it doesn’t solve the problem. And therefore, the approach we’ve taken is to start at the other end. Let’s look at the business problem that we’re trying to solve, rather than trying to solve the mess of data that organizations are typically trying to untangle. So we’re coming up with this now, and it’s a way of saying, “We know where our data is, we know what systems that sits in, we know how it flows through the organization, we’ve got a good understanding.”

VentureBeat: What is the business problem that you’re looking to solve with data?

Delooze: We’re looking at how we create insight for our members that we can then expose to them through the app. So you’ll see this through some of the challenger banks that will show you how you’ve spent your money. Well, that’s interesting — we can do that today. But it isn’t quite as interesting as a bit of insight that says, “If you actually want to hit your savings target for the holiday that you want next year, then perhaps you could do better if you didn’t spend it on these things.”

VentureBeat: What would that data-led strategy mean for your customers?

Delooze: If you can tie those two things together so that you can better understand your spending habits in order to be more successful in reaching your goals, that’s a great benefit for our members. So that’s something we’re really focused on at the moment, and we’re trying to build the capabilities to do two things. One is to create that insight, and the second is to then deliver that insight through the app and through the internet banking site that allows members to understand it.

VentureBeat: So how are you using technology to do that — do you build or buy?

Delooze: It’s a bit of a mixture. What we don’t want to do is to replicate Google BigQuery or Teradata or any of those kinds of tools. We already have quite a few of those. We have Teradata, we have QlikView — you name it, we’ve got it, basically. So for us, there’s two ways of trying to crack this problem — you can either try and hire the best team of data scientists and you build your own algorithms and, hopefully, two years later you come up with something that’s genuinely brilliant and unique. Or instead you go and talk to some organizations that say, “This is what we’re trying to achieve. How can your tools help us?” So we’re very much in the buyer rather than the [builder] category on this. We do recognize that it is our data, it’s our members’ data, so it has to be secure. The tools have to do what we want them to do. But I’d rather us focus on the higher-level value of building insight from the tools, as opposed to trying to build the tools and build the algorithms.

VentureBeat: Is that something you’re working on right now?

Delooze: Over the next two years, we’ll be building a new app and launching that to our members. Underneath that will be a much more data-rich or data-led insight view on members’ data and how they interact with us. And actually, at the moment, we’re just designing all of this; we’re re-imagining what the app will look like and how people will interact with it. We’re talking to our members about their experiences of how they work with us and others. We’re drawing on experiences from other organizations and bringing all that together and building that now. And then behind that, we’re building the data capabilities to create the insight to support that experience.

VentureBeat: What do you think the challenge will be?

Delooze: We have a very large and complex data estate. So that’s really about understanding what’s the best source of data and where do we draw that data from. For the last couple of years, we’ve spent some time building. We’ve changed our architecture, actually — we’re going away from this whole transaction-processing and batch-architecture model now to more of an event-based model.

VentureBeat: What technology are you using for this model?

Delooze: We use Apache Kafka to capture the events, and then we use MongoDB to hold the data from the event stores, and then we will use a series of different technologies beyond that in order to process that, analyze that, and serve that up. What we’re now doing is we’re trapping all the events that come through our architecture, so if someone makes a payment or sets up a direct debit, it’s an event. We track those events and we now feed those into a data store. So what we now do is we create a full picture of all the events that come through the organization. So if you’re a member, then we can see exactly how you’re transacting with us, through what channel, when, how, and at what time. And what we get is a full picture.

VentureBeat: What does capturing those events mean for the business?

Delooze: That’s allowing us to create a richer picture because we can augment that with data from the outside world. We can augment it with some of the ways in which members interact with us through the contact center or the branches, together with their digital footprint, and then you can start to see the full 360-degree view of the member. So for us, whilst we’ve got this big federated estate of transactional data stores, the real exciting area for me is that focus on events-based processing because I think that focus on events gives us a much richer set of insights than just the plain transactions.