Big data is big question for big banks

"Big data" is the catch-all phrase referring to the vast amount of information accessible through technology and too complex to process using traditional applications.

At an SAP financial services conference in London earlier this month, bankers from around the world explored the opportunities and risks offered by big data.

The ability to detect fraud, cross-sell to customers, gain loyalty and manage risks are some of the pluses. On the downside, banks are coming to grips with the ethics of using customer data and how to make the data meaningful.

"You don't just look at big data for [the sake of] looking at it; it's always to read something out of it, or to meet some requirement, so you need to make it meaningful," said Falk Rieker, global industry business unit head of banking at SAP, a global business management software solutions company.

SA banks have tagged big data as a key concern, according to PwC's latest banking survey.

Big data the new buzzword

"Big data is the buzzword for using data better to keep customers happy," said Johannes Grosskopf, banking and capital markets leader for PwC Africa

Banks are adopting a more holistic approach to customer relationships and using big data to identify customer needs and offer more client-specific products, according to the PwC report.

Simon Paris, global head of financial services at SAP, said big data was being used to build a customer-centric model for banking drawing on social media such as Facebook and Twitter, as well as analysing customers' spending patterns.

A UK bank, for example, analyses its customers' utility bills, looking at what a customer pays, how much they consume and how many people live in the household. It compares that with everyone else living in the same area. Based on this, the bank may then offer to switch the customer to another utility provider at a cheaper rate, said Paris.

"So you see this trend of leveraging big data to get much more personal financial management," he said.

The Commonwealth Bank of Australia tracks clients' account transaction histories for the past 12 months and flags any transaction that deviates significantly from the past history. If it appears, for example, that the client has been paid a bonus, the bank might contact them to make them a targeted offer on where to invest the money.

An initial pilot of this programme saw a 38% conversion rate and an increase in revenue and the bank retained funds, said Andy Hirst, senior director of industry marketing for SAP.

Fraud detection

One useful application of big data is in fraud detection.

Hirst said banks could use predictive algorithms and past data history to identify unusual payment patterns. This could be done quickly, in real time, to detect and stop suspicious transactions.

"The Association of British Insurers estimates that only around 30% of fraud in insurance claims is detected," said Norman Black, industry director for insurance solutions at SAP. "Now we have the potential to analyse the whole claims database in real time and look for patterns and find the 70% of fraud that's going undetected," he said.

One insurer used big data applications to analyse its claims and found that a leading indicator of fraud was a client who had changed addresses three times in the past six months, said Hirst.

This would have been impossible for a claims manager to monitor and could only be detected when the software discerned the mathematical pattern, he said.

"The power comes into play when you are looking for something you don't know and can't specify," he said.

But banks are still grappling with the ethical framework of mining big data.

"How ethical is it to intervene in your competitor's Twitter stream to use tweets about how unhappy their customers are with their bank accounts?" Hirst asked. "There are banks doing it, but they are not talking about it."

"Consumer discomfort with having their data analysed also needs to be managed, particularly given recent concerns raised by former US National Security Agency worker Edward Snowden about Internet surveillance," Hirst added.

But Rieker said banks were increasingly interested in how to grow their revenues by using their customer data.

"After all, who has more information about you than your primary bank?" he asked.

Source: Financial Mail via I-Net Bridge