Article

Fleet's new math: ATM + Web = Kiosk

Fleet Bank is struggling to define the future of its ATMs, and one man is defining that struggle. While he doesn't call his rebuilt machines kiosks, those in the kiosk industry will recognize the ATM's new features and applications.

March 27, 2002

Ford's horseless carriage, Edison's light bulb, Bell's telephone, Marconi's wireless-all

Fleet in the Spotlight Name: FleetBoston Financial Corp. Stock Symbol: FBF (NYSE) Assets: $212 billion Number of ATMs: 3,900 Web site:www.fleet.com

these inventions replaced the known with the great unknown. But before they were accepted, they were all regarded with suspicion.

Imagine being the person charged with convincing senior management to convert telegraph lines to telephone lines, or to retrofit kerosene lamp posts to electrically powered lamp posts. The job would call for tact, diplomacy, and a creative grasp of return on investment (ROI) possibilities.

ATMz-2-kiosks

Now imagine being chosen to guide how Fleet Bank revamps its fleet of about 3,900 ATMs, some of which have been operating for 15 years, with new Web technology. In other words, turning ATMs into kiosks that can sell tickets, pay bills, approve loans or perform any of a number of applications that can be added as desired. The job of changing a financial institution's thinking, not to mention infrastructure, requires tact, diplomacy and a creative grasp of ROI.

Jim D'Aprile is perfect for the job.

Ambassador to the Fleet

A 17-year veteran of the ATM market, D'Aprile began his career at the Bank of New England. He moved to Fleet in 1991 when it acquired his former employer. D'Aprile

Jim D'Aprile Age: 45 Title: Vice president of self service/ATM banking Current role: Architecting Fleet's self-service channel Years in the ATM industry: 17

Jim D'Aprile is moving Fleet's ATMs to deeper waters.

is now vice president of self service/ATM banking for FleetBank Boston.

In 1991 the world had not yet experienced the Internet. Now, D'Aprile, whose decisions will affect the success of the fifth largest ATM network in the U.S., is contending with a harsh reality-dispensing cash is no longer enough to bank on.

"The Internet is changing customer expectations," D'Aprile said.

At stake is how Fleet will develop new revenue for its aging ATMs and how it can maintain the loyalty of its 20 million customers by providing new features.

D'Aprile shows no sign of the stress of such high stakes. His experience shows in his resonant, bass voice, evenly modulated tones and carefully chosen words, qualities that make him an ideal ambassador to Fleet's executive management. His is the voice of reason and of experience. Even his second, functional title, and one he has held for three years, is modest: manager of technology research and planning.

As the architect of Fleet's self-service channel, D'Aprile will introduce Web technologies into Fleet's ATM machines, creating more choices for the financial giant's customers. Six full time staff members are working with NCR to develop the ATM application and with Zentropy Partners to design the user interface for the bank's rebuilt ATMs.

Not kiosks?

D'Aprile is careful not to call these machines kiosks, although they will be able to do much more than dispense cash, and although they will use Web components and have expanded self-service applications. The company's previous less-than-stellar rollout of five Internet kiosks probably makes it politic to avoid the term. D'Aprile wants Fleet's customers to feel that the new machines are as secure and familiar as the old. Hence, when asked if Fleet is changing its ATMs into kiosks, D'Aprile was discreetly indirect.

"Fleet is changing cash and balances to robust service and sales," he said.

He plans to test the new machines beginning late in 2001 to early 2002. Fleet will make its pilot machines available to employees in Boston. If all goes well, the new ATM with its Web components and expanded customer choices will roll out after the pilot.

Winds of change

D'Aprile said that Fleet must change its approach to ATMs because changes in the market are making cash machines less profitable each year. At April's KioskCom 2001, D'Aprile summarized for kiosk industry leaders the major problem for Fleet:

"Our network was suffering from basically the ultimate killer application gone too far: cash."

In a follow-up interview, he explained from his Boston office that while customers are still withdrawing cash, buying stamps and making deposits at ATM machines, growth has stopped and the market will probably start shrinking.

Between 1988-1997 ATM transaction growth averaged 12-15 percent. In 1998 growth had slowed to seven percent. By 2000, growth was flat. Now more people use point of sale (POS) machines, such as grocery store debit machines, than ATMs, D'Aprile said.

In the meantime, ATM competition has increased from traditional and non-traditional banking companies. Traditional banks such as Wells Fargo and Bank of America are already experimenting with Web-enabled ATMs. Non-traditional banking companies such as E*TRADE and American Express are setting up huge ATM networks. E*TRADE has set up an ATM network of about 9,500 ATMs as part of its plan to offer brokerage transactions as well as banking transactions.

American Express, a hybrid between a bank and an independent sales organization (ISO), has set up a network of more than 8,700 ATMs.

More touchscreens and more choices at the touchscreen are raising customer expectations. To be competitive in this market, Fleet must give its customers what they want. It's the old case of keeping up with the Joneses, and it's D'Aprile's job to know what that means to Fleet.

Fading technology

Some complications are caused by Fleet's growth. When it merged with Bank Boston, for example, it acquired Diebold ATMs that it now has to maintain in addition to its NCR systems.

"Our costs for development are going to go up," D'Aprile said.

Other complications arise as old ATM software becomes obsolete. Though no one seems to know exactly when, experts are certain that in the near future IBM will be ending support for its OS/2 operating system, the system used by most of Fleet's ATMs.

Replacing technology calls for careful design. The interface needs to be quick, as Fleet management doesn't want customers to wait in long lines. The user interface (UI) must accommodate a keypad and touchscreen rather than a mouse and a keyboard, and customers must find the interface to be intuitive.

D'Aprile knows that creating sweeping changes in a cautious banking company isn't easy.

Changing attitudes

Like any ambassador, D'Aprile is faced with political challenges. To negotiate his difficult task, D'Aprile must understand his environment. He defines Fleet as a process-driven organization. Therefore, he set up a process to justify changing direction. D'Aprile started with research.

He gathered his information through market analysis, focus groups held in Boston and New York, and a quantitative study Fleet commissioned. Both Dove Consulting and the Boston Research Group conducted studies.

The research results painted a picture of a financial industry moving to offer more self-service as part of a bank's overall services. Data also suggested applications Fleet might develop. D'Aprile learned that 21 percent of the people who use the phone to gather investment information would like to do so on an ATM, and that 35 percent of Fleet's customers wanted to pay bills on an ATM. These aren't staggering figures.

"Customers are not beating down the doors demanding expanded services. But we do see a desire for them and we expect that will grow," D'Aprile said.

Armed with facts, D'Aprile devised a strategy. Fleet would add Web capabilities to ATM machines in order to add more services to customers than was practical to add without the Web. Next he needed to take his message upstairs. He was fighting against history and misapprehension.

Overcoming the past

The history working against D'Aprile was Fleet's first try at a Web kiosk. Fleet set up five Internet kiosks in 1999 after D'Aprile pitched the idea in 1998. Though the kiosks are still running, executives didn't like them because customers didn't like them. In addition, the company didn't have a way to monitor how they were being used.

D'Aprile said that customers are more satisfied when the application includes new abilities along with what's familiar, but that the Internet kiosks offered features separate from what customers were used to.

"It was a non-integrated solution in a separate channel," D'Aprile said.

The new path to success lies in leveraging the current system, not ignoring it. It's a long path-D'Aprile pitched the new idea of Web-enabled ATMs to executive management in 2000 after four months of planning.

Bad choice of words

D'Aprile's idea of Web-enabled ATMs caused quite a reaction-all negative. The problem was his phrase, "Web-enabled."

"That was a big mistake," admitted D'Aprile.

D'Aprile casts management's reaction as, "What the heck is a Web-enabled ATM?" followed by, "What the heck is D'Aprile thinking about now?"

Unfortunately, the mental image evoked by his "Web-enabled ATM" was a long line of customers held up by a teenager surfing the Web or playing video games at Fleet's machines. Following his presentation, D'Aprile said he had to perform triage. He found that his four months of building a business case for Web-enabled ATMs had set him back four weeks.

"We had to dispel the notion of card holders playing video games in lobbies and convince executive management that it's not customers surfing the Web. We're going to replace the old portion of our platform with technology to leverage it," D'Aprile said.

Once D'Aprile redefined his plan, he had better luck.

"We were able to re-fix management focus away from the technology itself and on to its application," D'Aprile said.

Once explained that way, management understood the greater possibilities for customer satisfaction and services with Web technology.

"Things that you couldn't even consider before now become feasible, at least for consideration."

For example, Fleet could sell baseball tickets through its ATMs. Such an opportunity could expand the use of its ATMs beyond its own customers, giving the potential to create new customers. Selling baseball tickets might not be justified by a business plan, but for management, it represented the possibilities Web technology made available. In April, executive management approved the general idea to move ahead with putting the Web on ATMs.

Further application ideas Fleet is considering include other ticket sales, maps for tourism, and sales limited to specific locations of Fleet's network. Such limited scale applications aren't possible without Web architecture.

ATMs in the open

An ATM's specialized, often proprietary "states and screens" programming, where each screen contains one program state, such as "yes," "no" or "$10," makes it expensive to develop. Under those conditions, it's nearly impossible to build a business case for new applications or to argue for conducting a pilot. But open architecture makes it more feasible to add new applications.

"If you want to offer bill payment on an ATM and you already have bill payment via your online banking channel, if your ATM is Web-enabled you can re-use a significant portion of what you already have built," D'Aprile said.

"Open architecture means that if it doesn't work you can take it out," D'Aprile said.

He explained that legacy ATMs are the last holdout on OS/2 operating systems. These machines use software specific to certain vendors, such as NDC for NCR and CSS for Diebold, that are difficult if not impossible for anyone else to modify. On the other hand, Web architecture such as the networking protocol TCP/IP, Web languages HTML and XML, and capabilities of Microsoft's Windows NT operating system are independent of the vendor.

This independence allows you to deploy one application on an NCR, Diebold, Fujitsu or any other ATM. Open architecture also lowers development costs, D'Aprile explained.

The high cost of change

While D'Aprile didn't say how much Fleet will spend to make ATMs ready for Web-based applications, he cited an industry guideline of about $3,000 per ATM. With 3,900 ATMs, that's a potential cost of more than $11 million.

Just changing operating systems is expensive. D'Aprile chose Microsoft's Windows NT as the replacement operating system, as both Diebold and NCR are familiar with it. He discovered that it would cost $1.5 million to convert 1,000 ATMs to NT. D'Aprile was reminded of the Y2K bug crisis, when financial companies spent millions to fix systems just to stay in business doing the same thing.

In addition to operating systems, many ATMs need processor upgrades. Some machines have no processor at all, and a great number use older Intel 386 models, whose speed and computing power are too limited for Web applications.

In 2000, knowing that they would eventually approve some plan to add Web capabilities to ATMs, Fleet's management approved an infrastructure upgrade (management blessed D'Aprile's plan in April). Nearly 40 percent of the machines were upgraded to add a processor.

Even with these technological changes, D'Aprile pointed out that it's not the technology that's key.

"The bottom line of all this is to resist the temptation to deploy new technologies just because they are available," D'Aprile said.

The most important concern is giving customers what they want, in a way that's clearly defined as a business case.

"Find out what your customers want, what they'll use, determine what your business needs are. The technology is not the story - it's the application of the technology that is the story."

And James D'Aprile is the ideal ambassador to sell that story.