Having started as a traditional project finance company that had only a couple of product lines, Mumbai, India-based ICICI Bank was able to collect data in separate data marts that didn't communicate with one another without any impact on its operations. But, when the bank started adding product lines and began to emphasize cross-selling and customer acquisition, the separate data silos proved to be an obstacle, says M.S. Juneja, ICICI Bank's risk head, retail assets product group.
To gain a comprehensive view of customers and their risk profiles, ICICI Bank ($25 billion in assets) needed a data warehouse system that integrated customer data across channels and lines of businesses for holistic portfolio management. In early 2001, ICICI officials examined solutions from India-based companies, but the financial institution decided they would not meet the needs of a world-class bank. In stepped Teradata, a division of NCR (Dayton, Ohio), which was already working with some of the world's largest financial institutions and had a data warehouse system that could handle large volumes of data.
ICICI signed a contract with Teradata in late 2001. To accommodate the new platform, the bank migrated from a single node SMP WorldMark 4400 server to a dual node MPP WorldMark 4485 server with Teradata Database V2R4. The Teradata software runs on the vendor's proprietary server. "The data warehouse was installed in steps because the bank changed core banking systems three times" through 2003, Juneja explains.
Phase 1 of the data warehouse installation focused on building a single customer view. Phase 2 was aimed at using that information to understand the customer and involved integrating the Teradata data warehouse with a front-end tool for OLAP analysis.
To date, this information has been used to generate 242,975 leads, increasing the bank's cross-sell ratio in the process. Juneja credits 24 percent of the bank's recent sales to outbound calls from those leads. Those figures continued to improve as the bank implemented the third phase of the project, a campaign automation tool, which was completed in May.
ICICI used the data warehouse to generate 300 marketing campaigns in the past year, a number that Juneja expects to double in the next 12 months.
No Credit Checks
The data warehouse also serves as a de facto credit bureau. Since India doesn't have the advantage of the centralized credit bureaus that operate in the U.S., ICICI uses the data warehouse for credit scoring and credit-risk management. "Additional business is being generated by scoring transactions on current accounts to extend overdraft limits of good customers," Juneja relates. And, "Savings behavior is being scored to identify good-risk customers for cross-selling different products."
As a result of better prospect targeting and credit monitoring, between fiscal year 2003 and fiscal year 2004, the bank's retail deposits grew from $5.29 billion to $8.48 billion, and its customer base exploded from 6.6 million to 10 million, according to Juneja.
The next step, Juneja notes, will be more advanced data analytics, such as demographic clustering, stage-of-life analysis, propensity modeling and predictive modeling. "We want to build a data-driven culture across the organization [and] a best-of-breed business intelligence solutions team," Juneja says. "We want to make each customer profitable."
Institution: ICICI Bank (Mumbai, India).
Assets: $25 billion.
Business Challenge: Integrate data marts to gain a single customer view for cross-selling and credit-risk management.
Solution: Data warehouse from Teradata, a division of NCR (Dayton, Ohio).