Bas Geerdink, IT manager at ING Analytics, discusses how the bank is relying on Big Data technologies to improve customer interaction, marketing and fraud detection — with IoT adoption potentially just around the corner.
Banks are hundreds of years old. They sit on stacks of historical data on our purchases and our credit and debit history. They are the go-to place for our income, our mortgage and our loan applications. It’s to little surprise then that to date banks have formed the pillar of modern society.
And yet it could now be argued that today’s ‘legacy’ banks are under threat. Technology giants like Apple, Facebook and Samsung threaten to disintermediate banks from their customers, especially in payments, while Fintech start-ups and challenger banks could do the same with mobile banking and peer-to-peer lending.
These digitally-savvy companies are using technology to build improved user experiences, make transactions more seamless and affordable, as well as providing real-time updates on customers’ financial wellbeing. Simply put, banks are being outpaced.
It’s not just technology giants, and start-ups, that are changing the game though, with new legislation like PSD2 opening bank infrastructure up. Richard Muirhead, general partner of European software VC firm OpenOcean, touched on this perfectly:
“In short, post-PSD2 many banking customers will start looking at the bank as just infrastructure – a place their money is stored – whilst the app or company, maybe Facebook or Google, provides the old and new payments functionality under one shiny new umbrella.”
All of this makes life easier for customers, who now have more choice, but harder for those working in a financial services sector with tight revenues and increased competition.
Big Data, Big Challenge
Yet some banks are changing, and using newer technologies to do so.
ING IT manager Bas Geerdink, speaking at our Internet of Banking conference in London late last year, explained how the Dutch bank now sees itself as a “data-driven software company” through Big Data technologies, with the company starting to explore practical use cases of the Internet of Things.
ING has been a Big Data proponent for the last five years, investing millions of dollars into the relevant technologies and its data strategy, and restructuring so that it’s BI team (containing both the data warehouse team and analytics team) works within the customer department.
This last point is important because it shows ING’s focus in on using personalized or customized data to improve customer relationships.
In particular, Geerdink said that ING is looking at Big Data technologies to track the disparate data sources generated by their customers, even if this data is not owned by the bank and only accessible through APIs.
This is not an easy challenge, as managing more data sources, combining data streams, and – ultimately – taking both internal and external data and making sense of it is a complex task. Indeed, speaking at the same conference, Warwick Business School professor Markos Zachariadis argued that many banks sit on huge amounts of data, from which they could derive potential insights, but most are ‘not up to the task’.
In ING’s case, to bring this data together, the firm operates several data lakes and enterprise data warehouses, where static (batch) and streaming data are processed “for all kinds of purposes.”
In particular, the firm is currently moving to real-time forecasting on future expenses for the month ahead. One such example of this is ING’s mobile app, which went live last summer.
“The real-time account forecasting app provides customers with a forecast of 35 days ahead of their financial transactions,” said Geerdink. “The customer can use it to get a prediction about their future balance, which helps in short-term financial planning.”
The firm has also released Yolt, an app in the UK for people to “manage their personal finance, independent of the bank they are customer with”, a sign again of bank interoperability and the trend of predicting payments.
Improved marketing and security at ING
Away from the coalface, Geerdink says the firm is leveraging Big Data to improve both marketing and payment security, with machine-learning used on the back-end for batch processing and even fraud detection. With the latter, he said that outlier algorithms can predict irregular transactions, indicating that the card may be in the possession of a criminal.
“Batch processing is done with a variety of tools and techniques. In a way, the batch processes represent the core bank. Real-time / streaming data processing has taken flight recently, and the primary technologies used at ING are Kafka, Flink, Spark, and Tibco.”
Maintaining all this data is a constant challenge, but Geerdink says that it has a data lake for each country, with it working daily with the numerous open-source technologies, including Scala and Spark for predictive analytics.
Raw data is archived, while structured data goes to enterprise data warehouses, and is fed into a Hadoop cluster for ad-hoc analytics. Geerdink stresses the importance of these tools being able to deal with data ingestion, cleaning and blending.
Lambda architecture, designed to handle huge quantities of data through batch and stream processing, is used for delivering more customized mobile notifications, although Geerdink says that this is based on static data (and thus not real-time). Going forward, batch processing will enable customers to see stock market exchange rates, both historically and in real-time.
“They’ll see if stocks are dropping, or going sky-high,” said Geerdink. “We want to respond to these events and…to give the customer some real insights.”
Privacy in focus
Geerdink explains that the firm must be “very careful” as to what data it collects and when. Indeed, ING has been burned here before, in 2014, when it announced an experimental plan to provide customer payment information to third-party providers, so they could provide better ads.
As Computer Weekly explains, ING did say that this information would remain within the bank, and would be an opt-in choice for customers, but the bank ended up ditching plans after a matter of days.
“We have to be very careful. The first questions I get about this stuff [from management] is ‘what about privacy, security and ethics?’. You can’t believe how many legal people are involved.”
ING has discussed this at length on its own website around Big Data collection.
“The use of Big Data provides many opportunities, but also calls for caution,” reads a post. “We fully recognize that privacy is a very sensitive issue, and ING’s number one priority is the protection of our customers’ personal information.
“ING will therefore never give personal information to third parties that is traceable to the individual. Customers can rest assured that ING will only use their personal information if it is permitted to do so. We will always act in accordance with rules and regulations, as well as our own business principles.”
Watch our videos from the Internet of Banking below, including interviews with Monzo CEO Tom Blomfield, Secco CEO Chris Gledhill, Hellenic Bank head of digital banking Natasha Kyprianides, VC Spiros Margaris and many more! For more videos, visit IoB TV.
Banks can be “at the center of the Internet of Things”
ING’s Geerdink believes that banks should be leveraging the Internet of Things to build even better products and services.
Through processing sensory data, Big “static” data in batches, and using analytics engines and vizualisation tools and reactive software, he said IoT could be the next step to building even better products and services.
“We’re already looking at this from a strategic perspective for two years,” said Geerdink, adding that the bank already has an app for smartwatch owners.
“In a way, banks are already at the center of the IoT: they are connected to mobile phones, payment terminals and ATMs,” says Geerdink, further noting how micro payments can connect coffee machines, fridges and cars and payment terminals back to the bank.
“There’s a really good opportunity there. In the future, it’s going to open up even more opportunities.”
Banks, he says, could play two roles. “One is being the facilitator of the Internet of Things by offering micro transactions of APIs. The other is to be the provider by using data the customer already owns to build services on top of that.”
“The IoT will offer plenty of devices that can have a direct or indirect connection to a bank: cars, household equipment, manufacturing machines, etc. ING is always investing in new ways to make innovative products and services for our customers, and the connected devices in the IoT could be one of them.”
“Generally, banks are not yet adopting the IoT. Customers’ demands, rules and regulations are simply not ready yet to connect a highly secure bank to a great number of devices,” Geerdink said. “While there is a lot of talk and strategy, most banks are not yet connected to cars, household equipment, etc.”