Arun Mehta, global head of data analytics and engineering at NatWest Group has worked in the global banking industry for 25 years and, a week after his talk at Big Data London, was named Data and Analytics Leader of the Year at the 13th annual DataIQAwards.
Mehta kicked off his session at the Olympia-based show by emphasising how the focus on data within organisations has changed from collecting and storing it to analysing, monetising and making products out of it.
He added: “You can also now convert any problem into a data problem and then give the solution back to the customer. That’s how I’ve see the gravitas of data change.”
The NatWest data head has come up with his own recipe for data success, comprising of six key ingredients, which goes by the acronym SUPERB.
Superb data strategy
According to Mehta, the ‘S’ stands for Simplify. “The simplification of data, the simplification of the processes; the simplification of how you interact both internally and externally with the customers.
“For instance, if there are assets you already have in bank – ask how you can leverage them and maximise them. If you can’t, then throw them out. We’re currently introducing rules around data strategy that give us this kind of simplicity,” he said.
The ’U’ he explained, stands for ‘Understanding customers better’. He added: “You can have lots of technology and everything in cloud, but the key thing is what are you doing with these tech stacks? Hence understanding customer is very important.”
According to Mehta most of the organisations he’s worked for in the past, which include HSBC and California Bank & Trust, are really passionate about the customer experience. “They want to understand and predict customer behaviour. And the key to understanding customer’s data better is to make sure that it’s all integrated in one place,” he added.
’P’ said Mehta, is about protecting customers data. “There are laws and regulations that we have to honour so we need to ensure that whichever data assets we use we do so with the customer’s permission; the focus has to be on making our security infrastructure and architecture as secure as possible and encrypting securing data in the cloud.”
‘E’ is about engaging with the customers more effectively – for instance, doing the passwords by applying data science algorithms at the various data touchpoints that the bank already has. “It’s about investing in data science culture, investing in technologies and engaging our customers,” he said.
R stands for resiliency. In the pursuit of launching new technologies Mehta claimed that banks and organisations can easily lose focus on existing production systems. “Hence resiliency is really important. On top of functional requirements, banks need to focus on performance, security and scalability,” he said.
Mehta also wanted to add another ‘R’ here – the reskilling of people: “Bringing existing talent with you on your data journey – a set of people who are going to make it happen for you,” he advised.
Finally, ‘B ‘Mehta explained, stands for best data quality and best management practices. “Investing in that database so that we can have proper data linage.”
Mehta emphasised that while this was “not an exhaustive list”, if organisations focussed on these key areas, they would address most data problems.
Partnerships
Increasingly, banks and other organisations are leveraging data to gain customers in new markets, but another topic Mehta spoke passionately about at the data show was the need form partnerships with complimentary organisations that are established in the market they wish to enter.
“If you don’t’ have the datasets on customers in that geography, if you don’t’ have that historical data then the approach needs to be collaborated with companies that have a similar data and add it to the data you have to come up with a strategy that can work.
“In China, for example, fintech joint ventures with firms like [Chinese tech giant] Baidu, for instance, to gain better insights into the likings of the customers in that territory.”
According to Mehta, when it comes to innovation and launching new products, it’s a mistake is to follow the logical binary approach of either deciding to do it or not do it.
“What you need to do is to follow the use case approach – why are you are you introducing a new technology? And is it necessary that you need to put reporting on top of cloud from day one, for example. Lots of algorithms are compute hungry so you need to look and assess whether it’s necessary for the problems that you’re trying to solve now.”
He added: “It’s about creating cross functional teams with an objective – applying for seed funding and then, if it works, creating a robust case to get more funding so that we can mature the platform. We’re in the middle of that journey at NatWest, but there is a top-down backing for this.”
Data mesh
While the notion of understanding the customer, and pulling in data from a variety of sources, innovating quickly and decentralising into small domain-based teams seems to draw on some of the concepts involved in data mesh Mehta said that this methodology wasn’t something that organisations “could just do”.
He added: “There are building blocks – the way you go about federating your data, moving it away from centralised platforms. In our organisation we’re taking it one step at a time. We may not be doing data mesh in its entirety, but we’d like to decentralise the core data that we retain and give the power to the core business franchises we have. That’s a key part of the strategy we’re following.”
Share
