Balance of Power: Capital One’s Split-Focus IT Team

IT management involves two conflicting objectives, according to Roy Lowrance, chief technology officer of Capital One (McLean, Va., $46.3 billion in assets), a leading consumer lender and card issuer.

First, IT management has the objective of generating excess returns using new technology, which has a disruptive effect upon existing processes. Furthermore, attaining competitive advantage through technology often introduces new layers of complexity.

It's that very complexity that drives the other objective of technology management -- to control complexity by simplifying the IT architecture.

IT management requires leadership in both of these areas. "You need a curious blend of passion and rational thought," says Lowrance. "You need passion in order to find the ideas and generate enthusiasm around them, and then you need the rational thought to make sure it really makes sense to do it."

But in many banking organizations, the same team is responsible for both the passion and the rational thought. That can lead to a situation in which IT planners may hold back from introducing disruptive yet profitable innovations for the fear of complicating the system, or in which IT architects embark upon an oversimplification of the technology environment to the detriment of creative development.

To combat the possibility of internal conflict, Lowrance has established two groups within his IT organization, each of which reports directly to him.

On the leading edge, the "Strategic Technology" group seeks competitive advantage. "They look for game-changing new technologies to introduce, they get the business to support them, and then they introduce the new technologies," says Lowrance.

In the meantime, the "Global Architecture" group minds the store and makes sure that no idea is deployed before its time. "They're focused on making things consistent, making things standards-based, and getting reusable components deployed."

When it comes to reconciling the opposing camps, the decision comes to one place. "We have one executive at Capital One -- and that's me -- who has to deal with balancing these two," says Lowrance. "It's a good idea to have one person who worries about this, instead of having fifty.

"I'm the one who loses sleep over it," he adds. "We've found it to be an extremely strong way to approach this issue."

Putting Grid's Irons on the Fire

Grid computing is an example of the kind of technology discovered by the Strategic Computing group. In the first stage, back in late 2002, the group pondered the potential benefits of using pooled processing power to make better use of existing computer hardware.

Then, the group described the benefits of grid computing to its business constituents. "As the new ideas are surfacing in the concept stage, we reach out to the business," says Lowrance. "We invite both the business people as well as the other technical people to actually get briefed on the new technologies, and to work with us to try and figure out where they are applicable and how disruptive they're going to be to the installed base."

With the help of Capital One's bankers, the Strategic Computing group came up with a business case that it could present to the Global Architecture group. Based on positive-NPV (Net Present Value) calculations, the consensus was that grid computing had enough potential to warrant further examination.

But rather than rush into things, Capital One spent several months on testing the new technology in order to prove that it would work in its specific environment with its existing applications. Since the Global Architecture group hadn't been the originators of the grid computing idea, it became easier for them to undertake dispassionate tests.

Capital One entered the prototype phase in the start of 2003. "We took an application and rebuilt it using grid technology," says Lowrance. "We actually measured it, and also checked out all of the math to see if it's really true if we powered down one of the devices, the workload is recovered and sent to the other devices."

Fortunately, the technology worked as advertised. Grid computing passed the prototype phase and made it into the pilot phase, where it now supports a limited number of business units using live data. "We are deploying the technology in limited availability, with a lot of care and feeding to be sure that things don't go wrong," says Lowrance. "We'll be in pilot phase probably for about a year or so, and then we'll turn it loose."

The potential benefits are substantial. "Our paper and pencil work is that you can widely deploy this kind of grid and virtualization technology, and that it results in savings on the order of 25 to 30 percent of the infrastructure cost," says Lowrance. "It increases your infrastructure utilization rates, and this is a good thing."

Before joining Capital One, Lowrance was vice president, director and partner at The Boston Consulting Group, where he managed the entire IT practice area in the Americas. He has also held the post of CIO at Corio, an enterprise-software company; CIO at Fleet Financial Group; and vice president of corporate technology strategy at American Express.

For more on the Capital One grid implementation, read BS&T's June 2004 cover story, "What's So Great About Grid?"