The 21st Century ATM
presentation by North Communications
Banking Information Technology Secretariat
by Michael North
and Paul Kennedy
"It's been 25 years since ATMs were introduced, and consumers have embraced the concept of 'remote access' to their accounts for deposits and cash withdrawals, and the convenience it affords. The electronic banking center represents the next technological wave in which far more products and services are available..."
This paper briefly examines current trends in the financial services industry, specifically related to the Automated Teller Machine (ATM). It assesses the strategic forces in banking that have driven ATM adoption, and projects the potential effect of a new force -- the World Wide Web -- in the public access, self-service arena. The challenge to current Information Technology planning and implementation by banks is examined, and a new model, reversing many current paradigms, is suggested.
The Advanced ATM Defined
Today's ATM networks are connected to bank data centers through a diverse, complex set of network topologies and protocols. Some networks use standard dial-up lines; some use dedicated circuits; some are connected on a local-area network, if they're located in a bank branch. Some ATMs use network transports from IBM, NCR, DEC, Banyan or even variants of Internet Protocol. There are dozens of different configurations of buttons, prompts, menus and graphics, all running on ATM software that is mutually incompatible.
A powerful current trend is to migrate ATM networks to a single, universally-standard protocol -- TCP/IP. This is now being extended to the World Wide Web, the graphic segment of the Internet that attracts the most serious investment in new technologies, so that the user interface can be consistent across ATM families as well.
Much work remains to be done in the area of security, firewalls, interfaces to legacy systems, digital certificates, and encryption before all banks will feel completely comfortable using Web-standard software -- even on a protected, proprietary intranet not connected to the public Web. But that work, typified by the Secure Electronic Transaction (SET) standard now adopted by nearly every major credit card, bank and technology company in the world, is well-advanced, and the pace of development is picking up. We will soon have workable, industrial-strength bank networks; the time to plan, experiment and gain field experience is now.
The ATM of the near future will feature standard Web browser software, such as Microsoft Internet Explorer or Netscape Navigator. Under the hood will be the the current published software, modified by third-party overlays which remove the unnecessary, confusing (and non-secure) artifacts of the desktop PC metaphor. Fortunately, the base architecture of Windows 95 and NT enable this kind of overlay program, through the device of OLE and ActiveX. The Taskbar and Menubar, navigation buttons, scroll bars, alerts, checkboxes and lists, can be filtered, cloned, removed or represented on the screen as large buttons, easily touched on a touchscreen. The most challenging potential point of security access -- the keyboard -- is removed completely, sealed inside the ATM, and represented as a flexible keypad that can pop up on the screen any time.
Most important: the bank will have complete control over content. No idle surfing on the 21st Century ATM; wandering into inappropriate content will be impossible. The ATM will contain a continuously-updated list of authorized locations on the the Web (Uniform Resource Locators or URLs, the familiar http://www. address pointer), and the browser will politely refuse to go anywhere that's not on the list.
The advanced ATM will also be a multimedia ATM, using beautiful, natural motion video and stereo sound in multiple languages, smooth dimensional animation. It will be capable of live video teleconferencing with a remote teller at a video call center, using eyes and ears in a video camera. It will be able to scan, compress and transmit signed documents, printing MICR-encoded checks, send and receive email, read barcodes and smartcards, capture biometrics such as fingerprints and iris images, and print exact, full-size copies of any document. All of these features are available now on commercially-available products.
One Screen, Many Functions
"This is a whole new approach to ATM software architecture...a greater ability to accommodate new technologies. And because software applications are independent of the delivery channel, they can be leveraged across multiple channels, including information kiosks, telephone banking, personal computer banking and web sites."
All this means that the 21st Century ATM will do much more than dispense $20 bills and accept deposits. Alongside dedicated-function cash ATMs, and in new locations both inside and outside the bank, the ATM will soon offer a universal, free gateway to the world of electronic commerce. It will be possible to learn all about IRAs and college education plans, to find a job and the training needed for it, to file simple legal actions like divorce and small claims, to buy a book from Barnes & Noble or book a vacation through American Express, to check the sports replay on ESPN, to send flowers and a card to Mom, to buy and register for a cellular phone in all its intricate details, to send video, audio or text email to friends, to conduct targeted, incentive-based market research and real-time polling, to talk face-to-face with bank investment specialists...and much more...all from a free touchscreen at the local mall or bank branch.
No computer to buy, no baud rates to puzzle over or jumpers to set, no extra phone line or ISP subscription, no software to configure. The 90% majority of people who are still un-Webbed will have access to the People's Internet -- courtesy of their bank. What a pleasant surprise, when all they really hoped for in the past was a free checking account, if they were good customers.
Because every choice the consumer makes can be tracked and linked to his ATM card, a smart ATM will present a different face to each customer, with unique choices on the main screen. If you looked at a trailer for a Martin Scorcese film last month, the system can show you the trailer of his current release when it appears; if you sent a Bloomingdale's gift certificate to your mother for her birthday last September, the ATM can remind you and make compatible suggestions next August. The artificial intelligence and affinity-marketing databases, personal agents and so on to make this possible are already in use by cognoscenti (and clever marketers) on the Web today. The advanced ATM will put these powerful tools in the hands of banks, who will hold online auctions with content suppliers and advertisers for access to the bank's customers.
All this opportunity, and these exotic new functions, constitute a real challenge to traditional muscle-bound banking data systems. In the bank IT world, still very much dominated by the mainframe, many critical elements are unique, custom-developed and non-standard; banking data systems adapt slowly. To run an ATM network, of the advanced multimedia kind or not, many of the basic building blocks and decision points are the same; they include the ATM hardware and software itself; network protocols and telecommunications, server interfaces, databases, host interfaces, connections to financial clearinghouses and service bureaus, transaction management, internal accounting, central bank settlement, management reporting, marketing and customer service.
Change: Bottom-up or Top-down?
It is unnecessary to re-invent all these building blocks for ATMs to handle 21st Century electronic commerce. In fact, it is possible to simply change the front end -- the ATM itself -- and to let most other elements of the system continue to evolve on their interdependent tracks according to existing IT plans and priorities. This is a reversal of the intuitive top-down IT approach, that such a sea-change in function should require a complete re-engineering of systems, starting from the top of the hierarchy -- the mainframe -- on down. We recommend changing the lowest-order element first -- the ATM -- and then inserting a Web server into the hierarchy to control it.
This bottom-up approach implies the adoption of a new revenue-generating business model for ATMs, focused on building brand loyalty to the bank from the multiple services it now provides. In the past, transaction fees have been a negligible (and controversial) fee collected only from non-bank customers; ATM use has been free in order to encourage more use, because the business model has been centered on savings. In the future, ATM use should still be free -- except for transactions where service charges are customary and expected, like printing tickets. Revenue will also generated from advertisers and from website owners paying for screen-time and screen-touches, for printouts and client records.
Banks and ATM vendors should not attempt to modify existing ATM hardware and software to connect with the new world of ecommerce. Even if PC-based already, most current-generation ATMs are not up to the intense connectivity challenges and multimedia processing implied by the functions discussed above. They run a standard cash machine perfectly well, but they're not a general-purpose gateway to cyberspace. Trying to adapt them would be a lengthy, costly process; the results would be quickly obsolete; and smaller, more nimble competition, especially among the non-bank ATM network operators, would quickly leapfrog such an attempted retrofit.
Fortunately, a developmental firewall can be erected between the new generation of ATMs and current bank legacy systems, which are stable and must remain so, day in and day out. New ATM hardware and software is needed -- specifically, a powerful NT-based Pentium PC with at least 10 Gbytes of storage, MPEG video hardware, video conferencing connections, Ethernet and/or 56KB modem connection, and a flexible I/O expander to allow connection of many unusual devices, running a commercial Web browser with ATM overlay, all housed in a modular ATM frame flexible enough to reliably handle all the hardware options. Current network lines can likely remain in place, since TCP/IP routers and converters are available for every major network setup already.
A Web server, with appropriate protective firewalls facing in and out of the bank data center, needs to be added, but it does not need to perform any traditional bank data processing functions; indeed it is not desirable that it should. The ATM Web server's primary mission is to monitor, update, and gather data from the ATM network; it can interface with existing bank data processing software through the exchange of structured files, or through a transaction language like ODBC, which most bank software deployed in the last 10 years is set up to understand.
The Forces Driving ATMs
A quick recounting of the background of current ATM networks, and why they have been successful, is instructive in understanding why the move to the Web makes sense now.
When automated teller machines first appeared outside American banks in the mid-seventies, they were an unwelcome curiosity. The public wasn't interested in a new gadget they didn't trust. They were used to standing in line inside the bank to get their weekend cash and deposit their paycheck. Twenty-five years ago, ATMs were technological pioneers -- but like many pioneers they were unsuccessful at first, from a business standpoint. It took until the mid-eighties for the ATM industry to reach critical mass.
Three things happened in banking over since the early seventies that changed the landscape of financial services forever. And a fourth event is taking shape now...
First, economics -- despite the initial resistance from customers, some pioneering banks sustained their commitment to the ATM. They listened, learned and improved the machines themselves, because they understood that an automated transaction costs perhaps a tenth of a transaction handled manually in a branch. They sensed the tightening competition for cost, and knew they would have to close some branch offices, downsize staff, and consolidate functions in order to compete. The ATM was an important part of their re-engineering program.
Second, convenience -- consumers' lifestyles changed through the seventies and eighties: two-income, multi-job, 24-hour family became the norm. People traveled more on business, worked longer hours, were subjected to more stress and inconvenience. Suddenly the ATM looked like a convenience rather than a nuisance -- no more standing in lines, conforming to banker's hours, rushing through traffic to get to the bank before it closed on Friday.
Third, acceptance -- people got used to computers, first in very simple, subtle forms: in their car dashboard and engine, in their kitchen microwave, their VCR and video games. This trend accelerated as phones became voicemail gateways and menu-driven information utilities, and computers appeared on desktops, on the shop floor, at the checkout stand. So the initial mistrust of ATMs gradually dissolved away.
Economics, convenience and acceptance: that's why ATMs have become such a vital part of the financial services landscape. Now, flexibility delivered by the Web is a fourth trend that will strengthen each of those primary forces further. Web-connected ATMs are a new source of revenue for banks, a new convenience for consumers because they represent a one-stop shop; and they will be readily accepted, because they capitalize on the need for access to the Web (by both marketers and consumers) created by the drumbeat of billions of dollars of advertising money.
Considered this way, Web-connected ATMs don't seem so revolutionary; they are a natural extension of the forces that have driven adoption of ATMs over the past 25 years.
"What made Wal-Mart and Target household names-- one-stop shopping--will soon do the same for Automated Teller Machines. The ATM will take on a whole new meaning -- All Types of Merchandise. No longer will ATMs dispense just cash. Consumers will also be able to buy stamps, phone cards, travelers checks, even theater tickets, from a machine that has become a staple of convenience in society. Multi-service ATMs are becoming tremendous draws..."
So banks may one day be the consumer's gateway to the Web, a kind of People's Internet. In the process, banks have the opportunity to become the central hub of their customers' financial and information world, cementing profitable, lifelong relationships.
This broad liaison with today's banking customers, and with new customers like students who are just beginning their economic lives, can be initiated today at an ATM. It can then be extended through home computers, set-tops, WebTV, celphones and the whole universe of Web-aware devices now emerging. Forward-thinking banks have an opportunity today to secure this pivotal role, and to gain invaluable experience in the wide-open frontier of cyberspace.
to contact the author of this paper: email Michael North, email@example.com
Important websites: http://www.infonorth.com and http://www.kioskstore.com
Email for more information: VP, Sales and Marketing
at North Communications Los Angeles: Paul Kennedy, firstname.lastname@example.org;