Banks and real-time data – are they running out of time?

Published : 3 years ago, on 17/02/2022

By David Andrzejek, head of financial services, DataStax 

According to research by IDC Financial Insights, the market for consumer payments is changing massively – around 74 percent of all payments will go through non-traditional financial service providers by 2030. Gartner predicts that 50 percent of tier one banks will partner with fintech companies to deliver services by 2025. These trends are emblematic of the changes that banks face around delivering services to customers, now and in the future.

Why are these changes taking place? Because customers expect their banks to deliver financial services faster. For example, an investment trade used to take three days for settlement— T+3 processing, in security trading parlance — but in September 2017, settlement time shrunk to two days (T+2). Currently, the industry is working on T+1 timelines.

Similarly, retail banking service timelines are dropping from several days to minutes and even seconds. A credit card transaction can take place quickly, but it is only “authorized” at the terminal. The actual payment settlement will happen later, taking anywhere from one day to four to actually complete. Speeding this up relies on payment networks and standards, from Single Euro Payments Area (SEPA) in Europe through to the Federal Reserve’s FedNow service and the RTP Network from The Clearing House in the United States. These deliver instant payments where transactions are fully settled with funds available in seconds.

For retailers, this improves how quickly they can clear payments and have money transferred into their accounts. For customers, these payments can be made instantly, helping them manage their cash more effectively over time.

It’s time for real-time data

What does this all mean for bank IT teams? Implementing real time processes in banking requires that bank product leaders have to consider how they deliver those instant and personalized services to customers. This requires a real-time data strategy.

The move to instant payments is something that consumers and retailers benefit from. However, processing those payment requests quickly means that all the other payments-related processes have to take place in real time too. For instance, online credit checks for a loan need to happen in seconds. Fraud detection and anti-money laundering checks will need to happen even faster—in sub-second time.

As the research from IDC above states, there is more competition for the customer’s share of wallet around these kinds of services. To compete, banks have to shift from a product-centric approach over to a more customer-focused one. In practice, this means looking at personalisation across all the digital experiences that the bank offers. This covers both business services aimed at corporate treasurers and retail banking services for consumers.

Making banking services more personal

Modern banking services rely on data being used in context and in real time. Consumers today expect their bank to help them manage their money more efficiently and in their best interests. Offering services based on their spending habits and bills is one example – consumers will want to know if they will go overdrawn for the month due to their spending, so they can manage their account balances and avoid costs ahead of time.

This can be extended through smarter use of data before consumers place any orders so they don’t break their budgets, or to support immediate decisions on offering credit decisions or allowing the customer to use a buy now and pay later service. To achieve this, banks have to architect their data management and processing strategies in new ways.

Traditional banking IT deployments like core banking applications still run on batch processes, where all transactions are resolved overnight and reports are then generated to serve the business. Data can be used for other purposes, but this would rely on scheduled Extract Transform Load (ETL) operations being carried out to get the data to where it needed to be. This is another set of scheduled transactions that have to take place at specific times. This adds more time to complete each transaction that is just incompatible with modern customer expectations.

For bank product leaders, moving over to real time services involves re-engineering data infrastructure for “instant” processing and decision making. To achieve this, banks can build a “real-time data layer” over their core transactional systems. This data layer sits above the core banking system and has to capture transactions, move them to the appropriate other services, transform any data for other services, and then use analytics to make decisions on that data in sub-second time.

This system also has to accommodate a higher volume and velocity of data. As customers can access services from more places and over more channels, they will use those services more often. With mobile banking applications on users’ phones, transactions like checking account balances or requesting payments can take place at any time. This therefore makes availability another critical concern for that data layer.

Another consideration for bank product leaders when thinking about real-time services is how to make effective use of data for all the tasks that might be required. While simple services like a balance request can be provided, completing other tasks in short timeframes may mean looking at third party data as well as internal information sources. Examples here would be carrying out Know Your Customer (KYC) checks around opening up an account, or applying anti-fraud and security processes to a transaction.

Speed: The new differentiator

Macquarie Group is an Australian financial services company that provides banking services to customers. The team understood that, to differentiate their services, they would have to deliver what customers really want and achieve that at speed. This required a different data architecture, based on Apache Cassandra® that the bank’s team could take advantage of to build the kinds of services that other banks could not deliver. This approach allowed them to make the most of real-time data and build more engagement with their customers based on personalized services.

This real-time data strategy supports the bank’s overall modernisation and digital strategy, enabling their teams to deliver smarter digital services. This approach involves bringing together customer and third-party data and making it readily available to developers via modern APIs.

The bank’s product teams and developers can then use these APIs to build data-driven services for these “instant” applications. By using algorithms that cover fraud detection, credit approvals and payment processing, the bank can deliver services faster and more efficiently. This helps customers get what they really want from their bank – better insights about their spending and faster access to money.

Customer expectations of banks have continued to increase. We all expect services to be instant and always on, to know us inside out, and to have our best interests built in from the start. To deliver this, bank IT and product teams have to think beyond their core banking platforms and look at how to leverage data in real time. With less time to carry out those transactions – and the same stringent security, privacy and anti-fraud requirements to meet – this requires a new approach to real time data.