Bank of America: Three’s a Crowd?

Bank of America now stretches from the West Coast to the Gulf Coast to the East Coast. But its information systems don’t.

Even five years after combining with Southern banking giant NationsBank for $67 billion, customers still can’t easily transfer money between accounts from the bank’s East and West Coast operations. And now, Bank of America will try to absorb the 1,500 branches of FleetBoston Financial Corp., a New England banking giant that it is trying to acquire for $47 billion.

Bank of America has made some progress over the past five years in integrating its operations, but in effect it still has two core banking systems. Now it will have three. Experts such as Amir Hartman, a former technology consultant to Bank of America and founder of management consultant firm Mainstay Partners, says integrating those systems without angering customers won’t be easy, particularly if history is any guide.

When NationsBank (now Bank of America) acquired Barnett Banks of Florida in 1997, it angered clients with electronic glitches, poor service, and hikes in a wide range of fees. Result: Bank of America’s share of the Florida market fell 4 percent in three years.

Fleet Financial Group Inc.’s 1999 acquisition of BankBoston (to create FleetBoston) was a mess even by the company’s reckoning. Numerous branches were closed and BankBoston customers were force-fed Fleet products, leading to a revolt. Result: FleetBoston ranks last in customer service satisfaction among 120 banks serving the New England area, according to Consumer’s Checkbook, a non-profit organization based in Washington.

Bank of America customers complain that five years after the merger with NationsBank, they still have to jump through hoops to transfer money between accounts from the former two banks. Result: They wonder if this mega-merger will only make matters worse.

Marcia Salkin, a Bank of America customer in Washington, D.C., says she is frustrated at how long it has taken the bank to integrate its data systems. The California native chose Bank of America because she had a daughter moving East for college in Washington, and figured she might also be joining her a year or two later. By opening a Bank of America account for her daughter in California, she figured she could easily deposit money into her daughter’s account.

However, her daughter discovered she could not make a deposit to her California Bank of America account at a Washington-area Bank of America ATM machine– that required a separate trip to a teller. When Salkin did move East in early 2003, she discovered further problems. She opened a Bank of America account in Washington, but found that a check deposited to her daughter’s California-based account required a waiting period before the funds could be accessed.

And when Salkin went back West this summer for a visit, she too found that she couldn’t make a deposit at a California Bank of America ATM to her Washington account.

“The whole reason I chose Bank of America was because I thought I would be getting the benefits of a national system,” she says. “It would have been a lot simpler to sign up with a local bank.”

Eloise Hale, a spokeswoman for Bank of America, was surprised to hear of this difficulty and said it likely was due to special circumstances. However, other Bank of America customers who have had similar situations told Baseline of the same problems.

“They haven’t done a great job of integrating their systems to date. They’ve taken more of an ‘if it ain’t broke, don’t fix it’ approach,” says Jim Eckenrode, a retail-banking analyst with Tower Group in Needham, Mass. For the most part, Eckenrode says Bank of America has deployed middleware to create workarounds between the banks’ core systems, rather than replacing them. For the record, Eckenrode is a fan of this deal, based more on prospects for growth than on cost savings.

Behind-the-scenes technology integration usually doesn’t get much attention from Wall Street, but this time it is. Financial analysts were highly critical of the deal, in which Bank of America paid a 42 percent premium for FleetBoston’s shares. On conference calls and in investor presentations, Bank of America executives defended the premium, predicting they would wring out $1.1 billion in annual cost savings and generate another $200 million in “synergies” from operations.

Analysts say without the cost savings, the deal made little sense because FleetBoston operates in a mature market. However, studies show such levels of costs savings are rarely achieved in big mergers and acquisitions. A Mainstay Partners study of 450 large companies found that fewer than 20 percent achieved their targeted financial benefits from mergers or acquisitions.

To get it right, Bank of America will first have to sort out a few issues. For starters, it’s not clear what executive will be charged with the monumental task of merging and integrating hundreds of systems from the two operations. Halesays a decision will not likely be made on who will lead the company’s technology strategy until after the merger closes in mid-2004. Tim Arnoult leads Bank of America’s technology operations, overseeing 2,200 workers and an annual budget of $2.6 billion. Joseph Smialowski, FleetBoston’s vice chairman of technology and operations, has 1,700 technology workers and a budget of about $1 billion.

Executives at Bank of America and FleetBoston weren’t available for comment on the specifics of their systems.

Eckenrode says there is little overlap between the banks’ core computer systems. Both have a history of developing proprietary applications and maintain large development teams. Nevertheless, there are some areas ripe for consolidationfor instance, the two banks use a customer information system from Hogan Systems in Dallas and a consumer-loan generation application from American Management Systems of Fairfax, Va.

FleetBoston is known as a Unix/Sun Microsystems shop, and in recent years Bank of America has also relied on Sun for new application rollouts. It deployed a mix of Sun servers on Solaris 8.0 to run its new mortgage origination application.

While Bank of America has developed workarounds to integrate core systems, it has made progress on unifying operations on some fronts. It replaced two proprietary customer data warehouses in 2002 with one massive data warehouse based on technology from NCR Teradata. The warehouse now serves as a single repository for information on more than 28 million households and 2 million businesses.

FleetBoston has taken a more heavy-handed approach to integration. When it bought BankBoston in 1999, Fleet ripped out most of BankBoston’s infrastructure and replaced it with its own. It just completed a new network command center at its data processing facilities in Albany, N.Y., bringing all of its systems and network management operations under one roof.

Given the complexity and the scale of the task ahead, a similarly heavy-handed approach may be inevitable for the new Bank of America, says Robert Smith, former chief executive of Security Pacific Corp., which was acquired by Bank of America in 1992 for $4.2 billion.

“They try to present these things as a merger of equals, but they are not going to like one another in the end,” Smith says. “They can’t; they’re all fighting for positions.”

What You Should Do To Integrate Systems

Everyone doesn’t win in a merger. Employees will have to be fired and systems jettisoned. It’s best to make the choices quickly.

In the push to cut costs and integrate operations, companies often make moves that anger customers. Without customer retention, few acquisitions make sense.

Move to combine common applications between two companies first. In theory, it’ll be easier.

Merely bridging systems may not be enough long-term—especially if future mergers are on tap. Map out a strategy for the future with standard processes, applications and systems.When you have to integrate systems