On the weekend of May 18, Bank of America plans to complete a technology project that has spanned the country and will end in one of the bank's most complex markets: California.
The state is the home of the old BankAmerica, and the bank there still runs a version of the system it used before it was purchased by NationsBank to form Bank of America (BAC).
"We waited to do California last because of its size and complexity," says Mark Pipitone, a spokesman, in an email. "Now we are taking the final step … when completed, our entire franchise will be on the same platform."
California's conversion to Model Bank, which Bart Narter, a senior vice president at the market research firm Celent, describes as a highly customized version of Fidelity National Information Services' (FIS) Systematics, is "something that they've been putting off … it should have been done a long time ago."
B of A has updated the system over the years, says Pipitone; the bank's most recent conversions were the October 2011 switch of Washington and Idaho to the Model Bank system.
Cost was also major motivator for the conversion, Pipitone says. "Down the road, the bank expects cost savings as a result of the investment in these changes."
B of A is in the process of eliminating 30,000 positions as part of a multi-year efficiency program. In November, it began informing some technology and operations employees that they were among the people being laid off.
A core conversion is a risky prospect. Toronto-Dominion Bank (TD) experienced a failure in 2009 when it switched the systems of its freshly acquired Commerce of New Jersey.
"The core system is the heart of the bank," Narter says. "If the core deposit system goes down, everything comes to a screeching halt … [switching systems] is a high-risk move, but it's also a high-risk move to stay on a custom-built system."
Large banks often run separate systems for years after a merger — unifying technology is not as simple as throwing a switch.
For example, when First Union bought Wachovia in 2001 and took its name, it immediately relaunched a unified website, but the changes were only skin-deep. For the next four years, its online banking login page had two forms for its two sets of customers: former Wachovia customers who signed in with a username and password, and former First Union customers who used a PIN as a third credential.
The complexities of such technology overhauls run much deeper than what customers can see, says Steve Ledford, a partner at Novantas.
"Whenever you're talking about the core systems of the bank, you're not talking about a single program you can modify and change," he says. "This is going to be a massive project and one that you don't want to rush … for that reason you do a lot of testing, a lot of planning before you start flipping any switches."
It is typical for large-bank conversions to take years of planning and preparation, and the final transition of customers from the old system can take days or even weeks to complete, says Jost Hoppermann, a vice president with Forrester Research.
The timing depends "on the size of the bank and certainly also the impact [and] the scope of the transition," he says.
Such conversions are rare because they are an "enormous risk" and because the cost savings are not immediate, he says. "Such an endeavor is costly, and the return on investment is not something you can expect within … two years."