FinTechBig DataBanks need a single view of customer data to remain competitive: Feedzai

Banks need a single view of customer data to remain competitive: Feedzai

The siloed nature of traditional banks means they risk falling behind financial technology firms under PSD2, a Feedzai executive has warned.

The siloed nature of traditional banks means they risk falling behind financial technology firms under the revised Payment Services Directive (PSD2), a Feedzai executive has warned.  

“We will fairly rapidly see many large technology companies seeking to become financial services brands, just like large retailers have before them,” said Richard Harris, the senior vice president of sales international at AI fraud prevention firm, Feedzai.

“Then they can aggregate some other financial services providers underneath their brands. To remain competitive in a post-PSD2 world, traditional banks will need to preserve their brands and focus on customer experience.

“When you open a new bank account or ask for a loan, you want it to be as quick as opening a new email address with Google. You don’t want to have to wait three days for someone to call back and be told you must go to the branch,” Harris told GTNews.

However, as institutions seek to improve their customer experience, they are running up against a key challenge, which is that, often by necessity, they have become siloed, whether to create efficiencies or as a result of acquisitions.

“The consumer credit side of the business doesn’t necessarily talk to the corporate side of the business. The credit card team and the mortgage application team probably sit in a different office,” said Harris.

“The consumer credit side of the business doesn’t necessarily talk to the corporate side of the business. The credit card team and the mortgage application team probably sit in a different office”

Harris argued that processing power was expensive ten to fifteen years ago, meaning that the amount of data and communication a bank could use was limited. So until now, it’s never been a consideration for those organisations to be able to benefit from each other’s data.

Many organisations are struggling to create a single view of their customers with so much disconnected data, according to Feedzai.

A recent Aite survey of North American financial institutions found that only 10% are currently leveraging machine learning to orchestrate their customer authentication. And just 40% use cross-product and cross-channel data to make decisions.

Banks will need to integrate the data sources – internal and external – that all contain different parts of the one truth behind each transaction.

“Banks will need to know everything from how many accounts a customer has to when they last called customer services, and whether they made a complaint today,” said Harris.

“We are moving away from siloed data and towards 24/7 real-time banking that is ‘always on.’ And the banks and financial institutions that take steps to ‘connect the dots’ by design will be far better equipped to deal with their customers and these new competitive threats,” he said.

Google is already ahead of the banks here. “I was being shown adverts by Google for nappies three months before my son was born,” said Harris as an example.

“I was being shown adverts by Google for nappies three months before my son was born”

The flip side of customer intelligence is fraud prevention. As global security measures are increasingly improving, it’s getting harder to break into accounts. So fraudsters are conning consumers into making payments to them with false identities. If the account holder makes the payment themselves, then they will pass any security measures.

Harris said: “Fraudsters can still pretend to be the gas company that you haven’t paid your bill to. Banks are in the challenging position of trying to understand the intent behind the transaction, and identifying intent requires identifying thousands subtle fraud signals that are always changing, and then correlating those signals across data sources that don’t normally talk to each other.

“This all needs to happen in real-time so good customers can go about their business.”

“These are the challenges that humans alone cannot address. Fraudsters are evolving their tactics too fast, they’re leveraging advanced technology themselves, and they’re hiding in piles of data. It takes machine learning to detect their subtle, changing patterns,” he added.

Related Articles

Dealing data: Lessons from Facebook’s personal financial data move

Big Data Dealing data: Lessons from Facebook’s personal financial data move

1m Rory Donnelly
The quiet revolution of Cloud technology

Big Data The quiet revolution of Cloud technology

6m David Beach
What financial services can learn from the Uber breach when preparing for GDPR

Big Data What financial services can learn from the Uber breach when preparing for GDPR

7m David Beach
Sibos 2017 day two highlights: Brexit and banking, and why 'data is the new oil' in financial services

Big Data Sibos 2017 day two highlights: Brexit and banking, and why 'data is the new oil' in financial services

11m Alara Basul
When big data meets KYC on blockchain, there’s no place to hide

Big Data When big data meets KYC on blockchain, there’s no place to hide

12m David Poltorak
Deloitte and Cloudera tackle Ireland’s big data dilemma

Big Data Deloitte and Cloudera tackle Ireland’s big data dilemma

2y GTNews
Big Data: Hadoop, hype and hope

Asia Pacific Big Data: Hadoop, hype and hope

2y Richard Hartung
Unlocking the value in big data

Automation Unlocking the value in big data

2y Adam Richardson