As the capability of information technology increases, it follows that the power, content, and scope of banking products will grow as well. That's certainly true of payment systems, which have grown in sophistication and volume during the last forty years. There are more payment systems today than there once were, they can be lightening quick, and many provide global coverage. Recent research at Tower Group has led us to foresee an era when payment systems will fragment even more and become specialized to the needs of particular industries or types of transactions. Take the plastic card, for example. Now more than 30 years old, cards are incredibly successful. Yet, despite the billions of card transactions, the underlying data for the authorization and on the payment itself has not grown much and has remained remarkably homogeneous. But in response to the growth of technological capability, card-based payment systems will be expanded to handle more data and, at the same time, serve more diverse, specialized needs. The result will be a world where card accounts are segmented, not just from a marketing approach (gold cards, for example) or a credit approach (debit or credit) but in terms of functionality and the data collected. The evolving purchasing card is a noteworthy example. These cards can now be used by corporations' employees to make general business purchases of less than $25,000 potentially a $400 billion market that today has less than 2.5% of its volume on credit cards. Overall, the market for purchasing cards is growing at about 90% annually, compared with less than 15% for regular cards. But a purchasing card is more than a new market. It is also a new technology that needs additional authorization controls and reporting about purchases. For example, Visa and MasterCard's data capture for purchasing card transactions includes 34 characters Level 2 data, in industry parlance more than a transaction with a standard card. Nine characters are for sales tax and the remaining 25 are for customer-defined data, such as purchase order number. The associations' Level 3 data capture can include even more details, such as price per item. These extra data are input at the point of sale by the cardholder or sales clerk. Specialty point of sale systems that can capture the extra data do exist, but they are not yet widely used. Today, only about 10,000 commercial suppliers can handle the additional data. And even fewer can accept Level 3 data, which will probably evolve as part of the emerging electronic data interchange standards. But by the year 2000, we believe that 90% of the three million potential U.S. commercial suppliers will be able to capture and transmit Level 2 data. New purchasing card processing systems have been licensed by Visa and MasterCard at First Data Resources, Total System Services, Electronic Data Systems, First Bank System's Rocky Mountain BankCard System, and CoreStates Financial's Synapsys Inc. American Express has built its own new purchasing card system. We believe none of the large, self-processing bank issuers are likely to develop similar internal capabilities, at least initially. Citicorp, for example, the world's largest card issuer and largest self-processor, does its purchasing card processing at Total System. The technology needed by the merchant acquirer will most likely be in the network software, though it is possible that certain unique needs might be met through the hardware. Additional technology is needed at the purchasing card customer's end to receive the extra data, generate input for selective authorizations, and interface automatically with the general ledger. Colorado's ProCard is an example of a product that provides the customer with both the purchase cards themselves (issued by NationsBank, for example) and the appropriate software. Third-party processors (like EDS), Visa, and MasterCard have all produced less sophisticated software that handles specialized purchasing card data access and reporting. The point of the purchasing card is that it can handle customer-defined data, not just additional data. Therefore, it has the potential to handle extremely specialized needs. Fleet cards provide a good example of unique card payment needs. There are about 25 million fleet vehicles in this country driven by employees or contractors for work-related purposes. In the long-haul segment of the industry, trucking firms experience heavy turnover among drivers who can't be given a regular credit card (or cash) and who may also be on the road for months at a time. In addition, the cost of fuel to a long-haul trucking company is very significant. Around 1970, long-haul trucking firms began to turn to specialized payment systems that could carefully control, monitor, and report on the usage of fuel by drivers at the major truck stops. These payment systems also allow the driver to get cash or charge maintenance services. Nashville's Comdata is the leading company in this niche. Using a proprietary mainframe system, Comdata can control purchases by location, driver, trip, type of item, price paid, total amount, and other factors, and enable these transactions to be tailored to each of its trucking customers. Substantial extra data (such as odometer readings) must be input by the driver or sales clerks. Comdata also sells a specialized terminal, called Trendar, which is found at 1,200 of the highest volume truck stops. Comdata's system also helps control and implement price discounts that are often negotiated between the large trucking firms and large truck-stop chains. Local fleets constitute a separate market. About 20 million vehicles are owned by utilities, governments, delivery services, taxis, and so forth. They generally don't travel overnight, use less fuel per day, and fill up either at private terminals or at regular gas stations and convenience stores. A separate set of payment systems, such as Wright Express, exist to serve this market. Far fewer extra data are collected than in the long-haul market, but the authorization controls are still more stringent and the reporting more comprehensive than for regular credit cards. Like the long-haul market, the local fleet card programs have never been cobranded. But this is about to change and the results may be very interesting. In 1995, Visa approved a Visa Fleet program and International Automated of Naples, Fla., is set to issue the first Visa Fleet cards this year. International Automated is partially owned by Ceridian Corp., the parent of Comdata. The Visa cards sold by International Automated, with First Bank as the issuer, can be restricted by the customer to be accepted only at limited locations and for limited purposes. International Automated is in the process of arranging for the merchant processors of the approximately 180,000 U.S. gasoline outlets to take the Visa Fleet card. International Automated's proprietary software will then be placed within the processor's network and will control authorizations, data capture, and reporting. Visa Fleet transactions can then be routed either through Visa's standard authorization and settlement networks or direct from the merchant acquirer to First Bank and then to International Automated and the customer. Initially, the Visa Fleet card will function much like the bank's proprietary, non-cobranded card. However, it may be more acceptable to the market simply because of the Visa logo. With time, the Visa Fleet card should be usable for additional transactions, such as vehicle maintenance, and at additional merchant locations. So, the Visa Fleet card could eventually set the paradigm for an industry-specific branded purchasing card with unique controls, data, and systems. This is the beginning of a trend to augment a plain-vanilla bank card transaction with minimal amounts of data to create a much more variegated product that serves unique needs of customers. It's not unlike a consumer goods product being just one thing for years, but then finally becoming many varieties of product. One prime example is the evolution of Coca Cola into a number of specialty products like Diet Coke, Caffeine Free Diet
