Royal Bank Turbocharges Its Data Warehouse To Speed Collections And Control Costs

At many banks, consumers who have fallen behind on loan or mortgage payments can still get new credit cards.

At many banks, consumers who have fallen behind on loan or mortgage payments can still get new credit cards. That's because the banks have separate IT systems to manage each service, and managers can't readily see that the credit card applicants are delinquent on other debts.

Last year, Royal Bank of Canada went live with a new data integration and analysis system that bank managers expect will boost the efficiency of credit management and recovery operations. Using tools from Minneapolis-based nQuire Software, authorized employees can retrieve complete customer financial profiles, including their status on loans, mortgages, and credit cards. The bank is counting on the system to improve the efficiency of its collection operations, as well as to help spot delinquent customers.

The Toronto-based bank was already operating a data warehouse, based on IBM's DB2 database running on an IBM SP2 server, with more than 2 terabytes of highly detailed loan, mortgage, and Visa credit data from its 10 million-plus customers. The warehouse is updated daily.

But the bank hasn't made the best use of the data warehouse to improve its loan and credit collections or to identify high-risk loan customers, said Mohammed Rifaie, group manager of information resource management and data warehousing. "We have a manual and tedious process," he said.

Although the data warehouse generates some 100 reports that identify people who have fallen behind on their loan, mortgage, and credit-card payments, it takes four employees three days to compile them.

"And still, employees don't get all the data they need," Rifaie said. In addition, the data in the reports often isn't presented in a useful or intuitive format. And reports generated are either mailed or faxed to users, creating even more delays.

Royal Bank had considered building a separate data mart that would feed off of the main data warehouse with information specially prepared for credit management and bill collecting.

It had also examined other data-integration technologies, including IBM's Data Joiner and Information Builders' Enterprise Data Access. But Rifaie said those products didn't meet the bank's needs.

Earlier this year, it tested the nQuire Suite, which includes browser-based query tools and a server with data integration and transformation capabilities. The system lets users access information in a data warehouse, aggregate or summarize data, and retrieve it in a format defined by the user.

One key benefit of nQuire is cost. At about one-third of the expected cost of building a data mart, it was difficult to overlook, Rifaie said. He declined to disclose the cost of the nQuire project, including software and implementation. But list pricing for nQuire Suite starts at $125,000.

About 35 to 50 credit managers and collectors at Royal Bank's headquarters and in its three collections centers in Canada will use the nQuire system, which the bank has been installing and fine-tuning in recent months. By pulling historical data from the data warehouse-augmented with information from core banking systems-nQuire will speed collections and help identify customers who are falling behind on payments or are in danger of defaulting on their loans.

The system will provide more timely data, formatted to meet particular employees' needs. "It can create data transformations or aggregations on the fly," Rifaie said.

"And it will allow us to save three days and four bodies," he said, noting that the employees who now generate reports from the data warehouse will be reassigned to other tasks. The only downside is that implementing the software can be labor intensive because of the up-front data modeling and mapping chores that are required, said Rifaie.

If the nQuire technology works in this application, it could be useful for more general data integration purposes throughout the bank, pulling information from multiple databases to help managers in other bank operations gain a single view of their customers. For example, the bank tentatively plans to use nQuire to integrate data from the DB2 data warehouse with sales and marketing information from an NCR Teradata system.

---

Institution:Royal Bank of Canada

Assets:$294 billion (Cdn)

Business Challenge:Improve collection efficiency and identify delinquent accounts (loans, mortgages, credit cards).

Solution:nQuire Suite

Key Quote:"It can create data transformations or aggregations on the fly. And it will allow us to save three days and four bodies."