William Chenevich arrived at Visa in October 1993 with a pretty good idea about the importance of technology and his role in managing it.
He didn't know the half of it.
Within three months, Mr. Chenevich's boss and in-house champion, systems chief Roger Peirce, had departed Visa International for First Data Corp. And the global payments association had gained a new chief executive officer, Edmund Jensen, a commercial banker who admittedly had paid only passing attention to technology matters while serving on the Visa board.
Meanwhile, rapid technological changes and demands from member banks around the world were taxing Visa systems built for very different, more stable times. To some observers, the bank card industry's technological Cadillac was losing its edge, and MasterCard, with a computer network better suited to the increasingly fashionable client/server techniques, was looking more and more like the model of the future.
Visa needed new thinking, and fast. Mr. Chenevich, group executive vice president for payment systems, hasn't stopped running since.
Though he has barely begun the marathon, reporting now to Wesley Tallman, the president of products and information services, he has attracted a cheering section led by none other than Mr. Jensen, who has taken Visa's technology imperatives very much to heart.
"As CEO, I have to have confidence in Bill Chenevich and his people," Mr. Jensen said recently. "It is important to make business judgments about the capabilities we'll need for the future, and the key in my mind is the lack of flexibility in the central mainframe system."
To get the desired flexibility, Mr. Chenevich must engineer the radical shift toward client/server and away from the company's "central mainframe" heritage. Meanwhile, the existing transaction processing machine, one of the most sophisticated of its kind in any industry, must not skip a beat.
"VisaNet, the core processing system, is still growing at 20% to 25% a year," Mr. Chenevich said. "That kind of growth would pose a challenge to any business, and we have a lot to do to keep up with the processing requirements and volume growth." Because of its scope and efficiency, much of VisaNet won't give way to client/server any time soon.
At the same time, "our products are changing drastically," he said, ticking off just a few examples: the evolution from magnetic-stripe encoding to computer chips on cards; the home and remote banking revolution, being pushed by the Visa Interactive affiliate; the effort to make transactions secure over the Internet, in which Visa is cooperating with MasterCard and Microsoft Corp.; and the creation of an on-line, global information network for members using IBM's Prodigy platform, which will be the epitome of client/server.
"We have a few things on our plate," Mr. Chenevich deadpanned. "And while we have the technology, the marketplace pressures, and the new- product explosion all hitting us at the same time, the banks are pushing us on cost."
This means Mr. Chenevich and his vast systems organization won't be able to increase spending or headcount, though some jobs will turn over to bring in new skills. Visa's annual operating-expense growth has fallen close to zero, yet Mr. Jensen set out to triple "strategic expenditures" this year, to $76 million, which might seem to be coming out of Mr. Chenevich's hide.
But that has become par for the technology course in banking and other fields. The answers may be in emerging technologies like parallel processing and frame relay. For example, Mr. Chenevich said he is well on the way to "cutting what we pay to outside telecommunications carriers worldwide by 20%. On more than $100 million a year, that's a big number."
"It's the same situation senior managements were faced with when the information technology manager would come up and ask for $2 million for a new computer," said Mr. Jensen, speaking from his past experience as vice chairman at U.S. Bancorp in Portland, Ore.
"We have to take a new look at strategies and ask what we need to get to the future," Mr. Jensen said. "And we also have to talk about costs. They have to come down. The architecture Bill is planning will bring down costs and add the flexibility we need."
The pressure is nothing new to Mr. Chenevich, 51. He spent decade-long stints at Citicorp and Security Pacific Corp., which both prided themselves on operations and technology leadership. At the latter company before its 1992 merger with BankAmerica Corp., Mr. Chenevich rose to president of Security Pacific Automation Co., which won praise for running the back office as a profit center.
Just before joining San Francisco-based Visa, Mr. Chenevich was executive vice president and director of information services at H.F. Ahmanson & Co.'s Home Savings of America in Southern California.
Mr. Peirce, a former IBM executive and patron of Visa's mainframe legacy, said he had long admired Mr. Chenevich and found him perfectly suited to the strategic development role initially carved out for him.
Because he joined First Data, the transaction processing powerhouse that some at Visa view as a competitive threat, Mr. Peirce is no longer Mr. Popularity with his ex-colleagues. But Mr. Peirce did Mr. Jensen and Mr. Tallman a great recruiting favor.
Mr. Chenevich's extensive experience within financial institutions, and close exposure to Visa through its systems advisory board during his Security Pacific years, eased his way into the culture of a company owned by, and run for, banks.
"Technology is the biggest challenge at Visa - their umbrella architecture has to change fundamentally," said Stephen P. White of Dove Associates in Atlanta, a consultant who has worked with Visa, but not in technology. "Bill Chenevich understands that."
The move away from highly centralized processing "is a whole cultural change that can't be accomplished overnight, and it goes to the whole concept of how they deliver products and services to members," Mr. White said.
Visa executives and members alike "think highly" of Mr. Chenevich, Mr. White added, and find the client/server movement "a breath of fresh air," despite the need for new investments that "may cause some pain."
Mr. Chenevich has adroitly navigated internal political shoals, educating himself and others about client/server and what some have viewed as its disappointing economics.
"People got accustomed to mainframes, where 70% of the cost is hardware- related and 30% labor," he said. The smaller client/server computers are much cheaper, but the need for network administrators, system integrators, graphical interfaces, and software makes them labor-intensive. Hardware is down to 25% to 30% of the equation.
But the benefits of information access and flexibility are what client/server is all about, and the organization must change its ways. System requests that used to be handled through an autocratic, hierarchical process go before a more open and democratic steering committee. Mr. Chenevich has delegated much of the priority management to James Curran, an executive vice president he recently brought in from IBM.
"You won't see us launching three-year projects anymore like PS 2000," which reduced fraud and chargebacks but required widespread coordination that has become impractical amid the decentralization, Mr. Chenevich said. "There won't be a lot of money for a service enhancement; more likely it will be there for a reengineering."
"Anybody can come to us with requests, but strategic initiatives will get first priority, such as chip or stored-value cards," he continued. "Then come 'fix-it' requests, then region-specific funding and general enhancements.
"And then all of a sudden we might get a request to go into China, where they want their own payment system, which was unbudgeted, and we have to do that. It's a tough balancing act."
