ABN AMRO Boosts Voice Capacity

Seeking to improve call center effectiveness, ABN AMRO is using or has committed to voice technology from S1 for its mortgage subsidiaries, which together serve 1.3 million customers.

Atlanta-based S1 serves the financial industry, providing Internet, call centers, voice, ATM and other technologies. ABN AMRO's mortgage divisions increasingly rely on automated telephone systems to handle customer service calls, believing that a sophisticated voice system will free customer service reps to handle only the most complex customer needs.

In December, ABN AMRO Mortgage Group (AAMG) began using S1's vCSR Voice software, which runs automated phone systems. At AAMG it runs on an NT-based server, said Mike McBrien, general manager of S1's vCSR program.

ABN AMRO's Atlantic Mortgage subsidiary has used vCSR for three years, said Chuck Miller, first vice president of loan administration automation at Atlantic Mortgage. Other subsidiaries, including mortgage.com and Interfirst, will launch it later this year.

The technology provides enhanced self-service capabilities, Miller said. Among other capabilities, vCSR remembers details, so callers don't have to enter basic information repeatedly with every call. For example, if someone usually chooses a certain option, the system will present that option before others, saving the caller time.

vCSR can also complete transactions automatically-for example, promise-to-pay tasks-eliminating the need for human intervention. The product provides flexibility, such as allowing customers to choose among loss-mitigation options, such as deferred payments, when they're behind in payments. This eliminates the need for collectors to call in person. The software can also display text on a Web page to handle online transactions, Miller said.

The technology improves the bank's efficiency because it frees customer service reps to handle more complex issues, he said. In the first month of using vCSR, AAMG doubled its percentage of callers who self-serve.

Increased efficiency is crucial because the mortgage divisions have seen huge growth in recent years-mortgage.com's grown 1,000 percent in the last half of 2001, Miller said.

In January, AAMG was evaluating natural language capacity, another feature of S1's technology. It will probably begin using it later this year, Miller said.

With natural language, the software doesn't rely on a structured script. Basic systems require callers to say "one" or "two" to choose a task, McBrien explained. More complex ones include directed scripts, requiring callers to use certain phrases, such as, "transfer money from checking to savings." Natural language capability, however, allows callers to explain freely what they want to do. A bank can fine-tune the application to deal with unexpected responses, he added.

The technology understands the voice regardless of a caller's accent, MCBrien said. "You could say anything."

However, "it's a lot harder to design and build a natural language system," he noted, so banks are less likely to embrace such technology.