Bank of America's Data Mapping Adds $1 Per Share
By being smart about its network of branches and ATMs, Bank of America has reduced its annual expenses by $800 million while losing very few customers.
Cutting the operating costs of the ATMs and branches from $5.5 billion to $4.7 billion “works out to be about a buck in the share price,” said Jon Voorhees, senior vice president for retail distribution execution, in a presentation at Esri’s annual mapping conference.
In 2008 the bank had 6,151 banking centers; this summer that number dropped below 5,000, Voorhees added.
Since the bank started a faster development program 11 years ago, it has acquired and integrated more than 1,900 branches, built 900 new ones, closed 1,400, divested nearly 200, renovated 3,000, closed 6,000 ATM locations and added over 3,000 new remote ATMs.
“We have been busy.”
To cut costs while retaining customers, the bank uses information technology. It tracks every single transaction by customer, location, time and channel, and then stores the information for 60 months of history — more than 20 terabytes — stripped of all details that would identify individuals. In addition to seeing customer activity, the bank maintains information on the capacity of each branch or ATM, any peak time issues, costs of each location, its total revenues, annual sales, first year revenue sales to a new customer and a pipeline of any changes planned in the network, such as new ATMs or branches slated to close.
“I have a 20TB database that allows us to do things we are pretty sure no one else can do,” said Voorhees, adding that the bank has several patents on its analytical processes.
Citing a study by BAI, the retail banking organization, he said Bank of America branch efficiency, measured by expense to revenue, is 30 percent compared to 43 percent for its peers. Bank of America branches serve more than twice as many households and it has a 50 percent higher return per employee. The branches average $2 million in annual profit compared to $1 million at other banks, he added.
Jamie Dimon, CEO of JP Morgan Chase , has been quite public about his bank’s expansion efforts, said Voorhees. Six or nine months ago Dimon said he was confident that Chase’s new branches would generate $1 million in net contributions at maturity.
“I am closing stuff making a million; you can set higher expectations. Our deposits are bigger, our customer base is still the same, and we have lowered our costs significantly.”
That’s quite a change from the roll Bank of America had been on before 2008, which Voorhees described as an aggressive growth program.
“We were go, go go expansion.”
The bank added 510 branches in three years, hitting 6,150 stores, as he calls them, and 5,800 remote ATMs.
Then came the recession and revenue dropped by $20 billion. The bank had to reduce costs, fast.
“We were making a lot less money. But our expenses from all those investments we were making were going up fast. And so we had to flip our strategy to simplify the organization to better serve the customers that we have…We’d bought everyone else’s distribution strategy and in the last five to six yers we have been on a task to transform this to a 21st century network. We have to deal with 20 years of someone else’s mistakes.”
Although the bank has been at the forefront of rolling out new technology, like mobile banking, “two-thirds of our customers still go into a branch every month, so the branch isn’t going away.”
With an average age of 40 years, the branches were built for a different era and his team is looking at each one to decide “do I keep it, close it, move it, remodel it or add capacity.”
If they decide to remodel, the next questions are when, and then whether it needs a simple remodel for $1 million or a full remodel at $2.5 million.
The group, which consists of 100 staff members working from 75 different cities (80 percent from home offices) uses Esri maps to display much of the data used for its decisions.
“This year we will make 2,000 to 3,000 site decisions and every single one involves Esri maps,” Voorhees said.
Voorhees said he uses maps, with different layers of information, for almost all of his decisions.
If the air conditioning at a branch fails and replacing it would cost $750,000, in the past he would have said yes, replace it. Now he looks at the map and his data to see the branch’s value, usage and behaviors, when the lease is up, and decides whether he wants to replace the air conditioning or just patch it up. It doesn’t make sense to spend three-quarters of a million dollars on a new system for a branch that will shut down in a year, after all.
He showed a map of an area south of Palm Springs to demonstrate how data and mapping help decision-making by showing where customers live and where they transact.
Each blue dot is a customer, he explained and the shades of the map show their dependency on a specific center.
“The medium blue do half their physical transactions for a teller or ATM at Jackson Square, the dark blue use only Jackson Square.” The map suggests the bank has to offload some traffic because Jackson Square is nearing capacity. By seeing where the dark blues live, he can determine that the bank should locate an ATM 1.5 to 2 miles south.The map makes the case for not adding capacity close to Jackson Square but instead moving it closer to where the customers live.
“I have a map that I bring up every day in my office. My boss brings up the same map, and I can turn on layers that show us—where's my network today? What's everything I've done to it in the last two years, and what's everything I have planned in the next two years? Plus, what's the end state of what I want it to look like? And so when something comes up with a question about site X or community Y, I can go quickly look at it and determine…Does that thing fit with what we want to do? And so what used to maybe take me a week could take me five minutes nowadays.”
The bank has a 25-year plan for New York City and has identified every place it wants to put in a facility; now it is just a matter of waiting for the real estate to become available. In Manhattan the bank has invested $200 million which has incrementally generated $600 million in annual revenue.
“You have to be patient,” Voorhees said.
And have lots of good data.