Optimising the power of data for FSIs

This article is sponsored by Rackspace Technology and AWS.

Image courtesy of Markus Winkler on Unsplash.

In the age of rapid digitalisation, data-driven organisations are outpacing their peers.

Enterprises, therefore, cannot afford to be left behind with outdated tools and strategies. But what about a sector as heavily regulated as financial services? How can FSIs start their data journey and explore use cases, all the while remaining compliant with data regulations?

To discuss these matters, senior FSI executives gathered for a roundtable titled: “Unlocking the Potential of Data in Financial Services,” organised by Jicara Media, and hosted by Rackspace Technology and AWS.

- Advertisement -

According to Shwetank Sheel, Director, Data Services Sales – APJ at Rackspace Technology, customer behaviour is rapidly changing, especially when it comes to what they want to get out of their transactions with FSIs.

“The need for customers to have the same experience when they’re working with their financial sector partners, particularly banks and insurance companies, (with the experience) that they have when they’re working with Facebook, when they’re engaging in their social life — it has really driven a sea of change in expectations,” he observed.

“This has led to a lot of pressure, particularly on more traditional organisations, to use that increasing amount of data that has come up, but to do it in a way that meets regulatory expectations, that meets internal stakeholder expectations, and that drives transformation back into business,” Sheel continued.

Data woes

As organisations scale up their business, the amount of data at their disposal also multiplies. The question is, what to do with such massive volumes of data?

For a global investment bank, balancing data access and privacy regulations is a crucial concern.

“When it comes to getting access to data, structuring it is a big headache. Those are the challenges — access and data privacy. Naturally, we’re always going to have our hands tied. Based on various rules and regulations, and what I’m allowed to do, it’s always a challenge,” remarked an IT executive from the investment bank.

In the case of a large insurance company in Southeast Asia, it was a matter of data centralisation and decentralisation, depending on the situation.

The insurance company’s CIO shared that they digitise all documents, and as far as the source data is concerned, they have a centralised repository with a data warehouse and analytics that takes care of this.

“But with centralisation and decentralisation, it is always oscillating between these two. Initially, everything was decentralised. Now, gradually we started centralising the data. When the volume exploded, there were various other challenges. We are again now looking at decentralising it,” said the CIO.

As far as Singapore operations are concerned, the CIO said they are not using “that much data” as of now. Thus, they believe centralisation is the right approach for the insurance company.

Meanwhile, complexity is among the main data challenges faced by a large international bank based in Singapore, according to one of its senior IT decision-makers.

“One of the major concerns is the legacy data which are being contributed by various markets and clients in various forms and shapes,” the decision maker noted. “The standardisation of data is important, and data modelling is important, but how do we integrate the data in the multi-cloud infrastructure? Then how do we view the democratisation of the data to the client?”

For a Singapore-based securities firm, age-old industry practices are now clashing with privacy laws, particularly regarding data silos.

“We were not so keen on using data from the past,” said the securities firm’s enterprise architecture head. “There were silos and silos of data. With the tighter data regulations, we are now facing consent challenges, for example, going back to customers’ Google accounts as far back as when the regulatory framework was not agreed upon. Now, having to backtrack on a lot of that consent, getting that documented, and making sure that the content and consent are achieved, permeated throughout our legacy systems, which are also still in silos.”

Overcoming roadblocks

As to dealing with data, common among the concerns of FSIs is how to democratise access and adoption across departments.

Moody’s Analytics, which has worked with Rackspace Technology towards this end, shared how they went about their data journey.

“When Rackspace Technology jumped in with us, the first thing that they helped us do is, ‘How do we make the data presentable to somebody who’s actually going to be able to use it?’ That’s a big challenge for us. If you can make your presentation layer intuitive, you solve a lot of problems,” said Louis Chapman, Senior Director, Predictive Analytics, Product Strategy and Operations, Moody’s Analytics.

“People might think this is a documentation problem, (that) you need to have meetings to go over this, (that) you need to be able to create videos and spread them across your organisation — all that stuff is relatively true. But the bottom line is, the design of the presentation layer needs to be intuitive, and you need to be able to promote the right data forward. So, jumbles of data, collecting them for a very long time, and then synthesising that into something very simple, that’s what we are doing,” Chapman continued. 

According to Rackspace Technology’s Shwetank Sheel, the maturity of an organisation plays an integral part in data collaboration.

“Where the data organisation is well funded and well-understood, we will typically engage with them as an additional pair of hands. That’s primarily how we work with Moody’s. On the whole, they know what they want to achieve,” he said.

“When we’re working with organisations who are earlier in that stage, it’s usually because the business stakeholders have yet to see the value of the data. And in that case, we usually gauge their seriousness. I’d recommend that they start small and figure out the use cases. As we all know, there are use cases for the higher profile within the organisation, and it will differ by organisation. If it’s an insurance company with a big call centre problem, whether they’re having to stop the call centre, or even perhaps using the recorded analytics that they have to collect, to figure out how to improve their IVR, (that) might be a killer use case where they can really return value,” Sheel explained.

Meanwhile, the Spotify model has also proven effective for some of Rackspace Technology’s clients, noted Hemanta Banerjee, Rackspace Technology Vice President for Public Cloud Data Services.

“How do you get people outside of the normal reporting hierarchy to collaborate? They (a client) set up guilds and chapters around specific topics, (such as) Power BI and cataloguing around supply chains, food procurement, and so on. Then they were very aggressively measuring what is (the) adoption of analytics in each of those value chains,” he said.

“They did not go into a meeting without a dashboard, and everything was presented out of the dashboard. So, you’re forced top-down for certain kinds of behaviours, (and you) set the underlying organisational structures for people to start to collaborate,” Banerjee added.

Ultimately, the key to democratising data comes down to organisational buy-in, which most FSIs are currently struggling with.

“Within finance, there are a number of organisations that are still getting the organisational buy-in and starting small, figuring out how to get the data into a usable format, then through to the synthesis question and creation of that governance and visibility,” Sheel said. “This goes all the way through to when you’ve got the buy-in, and then it’s about delivering that will keep your CEO looking good for investing. At every stage of that journey, the question is how to keep the show on the road and moving forward.”

Sheel concluded that their hope is for companies to recognise the demonstrated value of data and make it more central to their organisation, leading to the allocation of budget towards larger, longer-term data projects.