Royal Bank of Scotland (RBS) has built an analytics neural network, capable of assessing risk across its business customers' supply chains.
The bank is entering the digital age with the use of data analytics to derive value on banking transactions.
In 2013, it created a Customer Solutions Group (CSG), tasked with providing insight to RBS executives and customer engagement teams.
CSG has access to data from approximately 1.5 million business customers and sees more than £7bn per day and £1.5tn (US$2.4tn) annual transactions.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Alan Grogan, chief analytics officer at Royal Bank of Scotland, said: "We want to use advanced analytics to free up the data."
To help customers develop better business strategies, the bank needed more insight into the bigger macroeconomic picture and the behaviour of its clients.
He said the analytics would support business banking customers, and help the bank's risk management process.
SQL Server Parallel Data warehouse powers RBS analytics
To handle multiple terabytes of data and complex queries more efficiently, RBS implemented Microsoft SQL Server 2012 Parallel Data Warehouse. By using the Microsoft software, RBS has been able to cut a typical four-hour query to less than 15 seconds.
RBS runs an analytics ecosystem with software from several suppliers. It uses Teradata for some functions and it also has Oracle. Initially, Alan Grogan, chief analytics officer at Royal Bank of Scotland, and his team tried working with existing database installations, but loading up to 50 terabytes of data from multiple sources and waiting hours for a query to finish became frustrating.
"We had one custom query last year that was probably the most complicated query the bank had ever run, and it took three days to execute," said Grogan.
For its new analytics system Grogan selected Microsoft. He said: "We went for Microsoft because it outperformed against our objectives."
The analytics engine is built using Microsoft SQL Server 2012 Parallel Data Warehouse (PDW) preinstalled and preconfigured on an HP AppSystem appliance. This makes the system easier to deploy, while also significantly cutting costs, according to Grogan. He said: "I knew that it would be easy for my team to transition from managing SQL Server databases to SQL Server 2012 PDW, and the solution cost about 85% less than products from other vendors.”
RBS was invited to speak at Gartner's Business Intelligence Summit on 10 March in London about the project. Grogan claimed the way RBS is using analytics is unique in banking.
"Every business has a supply chain. If you start linking payments together you end up with a neural network." In effect, by tracing transactions and following the trail of money, RBS is able to map out its business banking customers' supply chain.
Why is this important? "You can help customers understand their exposure to risks," Grogan said.
For instance, he said manufacturing is worth £160bn to the UK economy. "If a certain precious metal in your smartphone is sourced in another country, like Ukraine, [disruption of its supply] would infect the UK economy."
It is a situation that car manufacturers like General Motors (GM) faced following the Japanese Tsunami in 2011, Grogan said: "Semiconductors in satellite navigation systems and computer systems meant car makers like GM suffered delays in production due to the Tsunami."
Banks have previously never been involved in providing risk advice on a customer's supply chain. However, Grogan said: "As the UK's biggest business bank RBS has a lot of digital information and we can look at supply chain risks in near real time."
Analytics as a revenue driver
Grogan described this use of analytics as "Banking 2.0" where the bank deploys big data and analytics so that every decision is fact based. It allows the bank to turn relationships to partnerships, he said.
When asked whether RBS will build a business around this new analytics capability, Grogan said: "I don't want to be a data shop, I am a bank, but I want to help customers meet their needs. Hopefully the risks to the bank will fall."
By using data analytics to put customers on a firmer footing, Grogan said the bank gains better insight into the risks associated with lending to a business customer. He said: "The bank will be on a stable footing and hopefully RBS will make money."
More articles on banking IT
He believes data monetisation is a decision that will need to be made with the bank's customers. After all, it is their transactional data that feeds the analytics engine. There are certainly organisations that would charge for near real-time suppy chain risk exposure data. But, Grogan believes the more the customer does business with RBS, the greater value the bank gains.
He said: "Personally I'm against charging right now. We want to lend to people but we also want our money back." And this is the real value the analytics provides to RBS.