Extract's Table of Contents:
In 1992, electronic commerce meant EDI, corporate cash management and banking funds transfer systems. In 1997, it means all of those plus the Internet and intranets. (Intranets are networks that employ Internet technology but that restrict access to information and services to employees and privileged trading partners.) For some commentators, electronic commerce and the Internet are equivalent in that they see it as the base for a massive consumer market for on-line retailing, selling of new information services and handling of payments, extrapolating from its recent 100% or greater growth per year in the number of users and providers of services on the World Wide Web. Forecasts of the size of the Internet market vary widely, even wildly.
"the on-line market is expected to reach over $1.4 billion by the year
2000" (Margo Komenan, Electronic Marketing, page xxiii) versus
"Forrester Research's estimate that the Internet will be a $6 billion market by 2000 and the firm prediction that it will be a $150 billion market" (IDC Research).
If the market turns out to be in the range of hundreds of billions of dollars rather than tens of billions, then obviously Internet and electronic commerce will indeed be synonymous. (Keep in mind that a billion dollars is just four dollars per household. The sales of Barbie dolls is more than $1 billion a year; electronic commerce on the Internet in 1996 was around half that.) But even if, as some experts predict, the Internet growth surge slows and its slowness and unreliability lead to a literal turn off, the mainstream of electronic commerce will still continue along its established path for business-to-business relationships. Its growth rate of around 20% a year has been sustained for a decade without the Internet being an important catalyst or tool. Obviously, the more the Internet is used, the more it will be used for electronic commerce. It extends the reach of value added telecommunications networks that are the main vehicles for EDI, make it simple and economical for small firms to implement most but not yet all the components of electronic commerce that large ones routinely employ and, perhaps most important for the evolution of electronic commerce, its technology is the base for intranets.
Our own view is that the Internet will not reach the hundreds of billions of dollars level in person-to-computer electronic commerce for many years. Its growth will be blocked in the short-term by its well-known problems of reliability and capacity, unrepresentative demographics - affluent technical professional males dominate its use - and peoples' being in general content with fax for communication, paper catalogs for mail order, 1-800 numbers for ordering, and printed newspapers and magazines. There are a number of striking successes in Internet retailing and plenty of activity in financial services, especially in ensuring secure use of credit cards. But the successes are largely confined to niche selling of books (Amazon), wines (Virtual Vineyards), computer hardware and software (many companies) and travel. The payment systems are largely in the pilot stage in mid-1997 and no industry leader has emerged. Of course it's possible that in the next few years users of personal computers who browse the Internet for information and entertainment will make it the home equivalent of ATM machines and 1-800 numbers and open up a mass market for person-to-computer electronic commerce, but whatever the optimists predict, their forecasts have no realistic base of evidence today. The promised mass consumer market has not yet appeared and there is no evidence that it will do so soon.
The Internet will, in our own view, have most immediate impact in business-to-business electronic commerce. If 1995 was the year of the Internet, 1996 was that of the intranet. Intranets use Internet technology to solve several of the longest-standing and most difficult problems in corporate use of computers and telecommunications:
- Reach: Who can access the firm's on-line services and information resources.
- Range: What information and services can be automatically and immediately
crosslinked.
- Ease of use and training: how natural is the system to access and navigate through its menus and options.
The Internet is in many ways a clumsy, insecure, and unreliable tool, but its weaknesses count little compared to its strengths in these three areas.
Reach: TCP/IP as the base for a new everything in telecommunications
Throughout the over forty year history of computing and data communications, the fundamental single problem has been that any computer system trying to link to another one has had to know what type of device it is. The main reasons for this were that the technology evolved by companies such as IBM, AT&T, Digital Equipment, Wang, Microsoft, Apple and the other pacesetters of their era developing what are termed "proprietary" (vendor- or technology-specific) systems and the "standards" (rules for formats and connections) and "protocols" they use. For instance, an IBM personal computer could interconnect to ones that used Apple's software and local area telecommunications networks. Little by little, first IBM's and then Microsoft's systems achieved such a dominance in the marketplace that other vendors either incorporated them or added facilities for interconnecting to them. Committees from the international telecommunications phone service monopolies took the lead in defining a massive blueprint for interconnecting everything to everything, called the Open Systems Interconnection reference model that provided a guide for "open" versus proprietary systems but it did not offer specific products. Those were slow to emerge and as the pace of technology innovation accelerated, with personal computers, multimedia and new modes of telecommunications transmission making this OSI model obsolescent, the problem of incompatibilities among systems.
Once telecommunications became cheap to install and operate, TCP/IP - Transmission Control Protocol/Internet Protocol - entirely changed all this. It's the bedrock of the Internet, a communications protocol that allows any computer of any type anywhere in the world to link to another computer of any type anywhere in the world. It's a clumsy and inefficient protocol; firms like IBM and AT&T had to build their services with security, cost and efficiency as priorities. Even the most simple of telecommunications applications, electronic mail, was marked by proprietary systems. To send a message by, say, MCI Mail, to a computer using IBM's Profs or Digital Equipment's Decnet was impossible without expensive computer hardware and software designed to make the translations in message formats and communications protocols.
The community of technical professionals, mainly in universities, who worked on or near what became the Internet took a pragmatic view of telecommunications - Connect and communicate and don't worry about efficiency, security and network management controls. They invented or enabled many innovations in telecommunications in order to make this first possible and then practical: packet-switching - ways of breaking telecommunications messages into small units that could be pumped through a network as an assembly line instead of tying up the communication channel in the way a phone call does - routers - the traffic managers of telecommunications traffic across dissimilar networks - and the World Wide Web, the key to turning the Internet from a network that only people with sophisticated understanding of information technology could use into something almost as easy as a VCR.
The transition took time. For several decades, TCP/IP and the software operating system called UNIX remained an exotic niche. ARPANET, Bitnet and NSFNET, the forerunners of the Internet were horribly difficult to use for other than computer-fluent experts. In business, TCP/IP was used only in UNIX applications on workstations. The mainstream of IBM and Apple personal computers did not employ it until the early 1990s. With the invention of the World Wide Web in 1992 and the rapid development of easy to use Web browser software, the strengths of TCP/IP came to the fore. It made the Internet a world wide service that, first, any personal computer software and hardware vendor could tailor its products to without having invest in complex new software, made it easy for personal computer users to access as either a provider or user of services, information and goods, and made it possible for companies to transfer information from their transaction processing systems built on IBM's and others' systems built on proprietary technology. On-line information services like America Online were both able and increasingly driven by the exploding growth of the World Wide Web to add "gateways" - computers that link to the Internet - to the degree that in practical terms there's now no real boundary between them and the Net.
TCP/IP is now the protocol of electronic commerce. That's not to say that it's the protocol of electronic commerce providers, only that when they use computer hardware and software to convert their telecommunications traffic from and to proprietary protocols, they have the reach for any-to-any communication - any computer to any computer. Computers now can link to each other as easily as telephones and postal services do. Trading partners have relied on phones and mail for most of this century in their person-to-person. TCP/IP is the phone/mail/fax of the 1990s.
Range: The genius of browsers
TCP/IP might have remained the way nerds and academics communicated with each other without the World Wide Web. Until 1992, the Internet was a network that could be used only if you knew how to type a complex set of commands to invoke its services. The Web defined a subset of the Net that uses what is called hypertext (Web addresses routinely begin with
"http" for hypertext transfer protocol) to connect information items - Web pages - to each other. Hypertext is a concept that dates from the 1970s that views information as a dynamic set of pointers rather than the tidy sequence of books and structures of data bases.
The Web provided an entirely new way of interacting with information on-line. Any Web page could highlight a link to another page so that, starting from a simple screen display of a Web site's "home" page, users of personal computers could explore a universe of hypertext links. The Hypertext Markup Language for creating Web pages and hypertext links could be learnt quickly and made it cheap to build basic Web sites.
It was the Web browser, though, that went beyond just displaying Web sites. The term is now somewhat misleading, in that the leading browsers, Netscape's Navigator, Microsoft's Internet Explorer and those offered by on-line services like America Online, are powerful tools for handling just about any Internet application, including file transfers and electronic mail, for building Internet applications and for ensuring security. While their initial purpose was simply to make the Internet easy to access and use, they solved the biggest single problem firms have faced in their use of information technology, one that we term Range. (For a detailed discussion of Reach and Range, see Peter Keen's
Shaping the Future: Business Design Through Information
Technology, Harvard Business School Press, 1990)
Whereas Reach defines who can access the firm's information and on-line services, Reach is the extent to which these are "integrated" instead of being separate systems that cannot be combined. The basic information and transaction processing systems in most organizations were built as separate applications, so that, for instance, banks used different systems and customer identification codes for handling checking accounts, mortgages, credit cards and savings. In pharmaceutical firms, clinical trial data in individual countries were built on different software, hardware, telecommunications and data base platforms. In most manufacturers, engineering, accounting, inventory and financial data were separate departments, systems and technology base.
There were many forces driving this lack of integration the limitations of the available technology, the rush to implement urgently needed basic transaction systems on a case-by-case basis, competition between software and hardware providers that offered very different features based on very different standards and technology foundations, and the dominance of the departmental structure of organizations in the 1960s through 1980s. During the '80s, the entire thrust of information technology supply and demand open systems, interconnection, and common standards. Electronic commerce benefited from this and electronic data interchange was one of the strongest catalysts for it - EC is impossible without standards for companies' systems to link to each other and share information.
Even with EDI standards, though, an almost intractable problem was that the information in far too many old transaction processing systems could not be accessed or shared easily. In many areas of business, firms did not have the technical staff resources, business justification, management commitment or available tools for developing accessible information resources. This was especially true in human resources, customer service in organizations where systems had evolved as "islands" of automation, research and development and public relations; these fell outside the mainstream of information and transaction systems.
Personal computers had, if anything, led to less not more information being accessible as a shared resource. They encouraged the development of even more local and departmental systems, with little coordination and in many instances with an explicit resistance to oversight from central corporate information services groups who were seen, sometimes accurately, as bureaucratic control freaks who disliked the intrusion of "amateurs" with "toy" computers. Apple Computer deliberately made its telecommunications and software incompatible with IBM's mainstream systems, a move that both meant that its products could not become the front-end "client" that accessed information and communication "servers" and basically lost Apple the big business marketplace.
Integration became the watchword of business use of information technology in the 1990s: integrated financial software, integrated customer data bases, integrated logistics and the like. The growing focus on business processes, teams and customer service as the building blocks of new organizational forms demanded integration of systems. "Legacy" systems blocked that and still does in many instances. The tools of the Internet brought a solution. Browser software plus the toolkit of the Internet programmer/webmaster/developer offered a way to make information look integrated and a vehicle that is so simple to use that for the first time in the history of information technology there is no need for training.
The key trick of the trade here is hypertext. The challenge for systems developers in integrating information across legacy systems is that they have to write complex software links, using what are termed client/server systems, application program interfaces, distributed relational data bases, agents, applets and other tools. Most of these would work only for given types of information using given vendors' software and data base management systems. TCP/IP solved the problem of how to move the information; it extends Reach but does not in itself provide Range. A browser like Netscape lets any type of information in any application or any data base be treated as just another Web page to be linked to. It's inefficient and can involve some complicated technical minutiae but it works. Here are two examples of what this means in practice:
- One of us is working with a phone company in Latin America, which has recently lost its monopoly and faces aggressive competition. One of its priorities is to improve its billing and customer service information capabilities. Large customers want consolidated billing, fast response to queries and a single point of contact for problems. The phone company has multiple systems that are hard to access, totally lacking links between them and not available to customers on-line. It's involved in a massive development project involving a number of European and other phone companies to develop a system that will cost it around $100 million. The project is well behind schedule and won't be in place for several years. Meanwhile, the firm is losing market share as national and US companies target its best corporate clients.
- Using Netscape Navigator and software development tools designed for building on-line applications quickly, the firm has an intranet system that offers consolidated billing and customer relationship data, built in a matter of months. It can't meet the requirements of on-line transaction processing. It's a little slow. But it works, It's easy to access and use. It cost thousands, not millions, of dollars. And it provides the base for the firm being able to develop simple systems in response to business needs in the same development time frame as the business's development time frame.
- One of the world's three largest pharmaceutical firms spent several years trying to find a way of integrating the worldwide information on clinical trials of new drugs without having to spend the GNP of a small country to do so. For decades, the US Federal Drug Administration would not accept data from other countries. There was thus no incentive to integrate the systems that processed data in the 160 countries the firm operates in. Now, the FDA allows and even encourages the submission of data from anywhere, in its concern to speed up time-to-market for life-saving drugs, especially for treatment of AIDS.
- The firm's head of information systems told one of the authors of this guide that the lowest price the top systems integration firms like EDS and Andersen Consulting thought it would cost to build a worldwide clinical trials system was $200 million. A year later, the firm has one, built in about three months by two people. Again, it's slow, incomplete, can't handle transaction processing. It works, though and it's easy to access and use. It is a front end to the existing incompatible systems across the world. Netscape Navigator is the window into what looks like an integrated system. The Web is the access vehicle and TCP/IP the electronic messenger service.
Ease of use: Hypertext as point and go
Navigator, like Internet Explorer, Spyglass, Mosaic and other browsers also offers a deceptively simple feature that is as radical in its impacts on the use of computers as spreadsheets were for personal computers: we term it mundaneness. "Mundane" means that you take something for granted. Think of renting a car at an airport. You don't need to look up a manual to get the car started. You probably can't even recall exactly where the instruments were on the last one your used.
Computers have not been mundane. The dreadful term "user-friendly" was thrown around in ads touting a new piece of software for a full decade. It really meant "less user-hostile than last year." Computers are still difficult to use. The genius of the combination of browsers, the Web and hypertext is that they make intranets simpler to use than any other information retrieval system. To our knowledge, there's no "Learning to Use The Web" training industry. Navigator offers a commonsense, natural - mundane - way to use computers to find information, ask questions and make transactions. Its users don't need to know anything about the type or location of the many computers that may store the information that can be accessed from a single Web page (TCP/IP handles that), its nature (hypertext links make numbers, pictures and videos just "objects" to be transferred) or the software procedures for obtaining it (the browser disguises the links in the same way that the phone disguises the complexities of voltages, cycles, "hand shakes", error detection, call accounting, latency and hundreds of other technical issues; you just pick up the phone, dial the country code number and say "Hello."
Internet browsers aren't quite that mundane but they are still by far the most natural and easy to use software available on personal computers- and most easy to enjoy using.
Summary: Internet-enabled electronic commerce
Clearly, no firm can ignore the Internet in its planning for electronic commerce. Clearly, no one - including us - can predict the future of the Net and its impact on big business, small business` politics, education, government, entertainment, law, banking... - life on this planet as we know it or can imagine it. Both of the writers of this guide are enthusiastic skeptics about the Internet. We distrust hype. We have both worked in the field of electronic commerce for well over ten years. We've been enthusiasts for years. Our enthusiasm predates the Internet. Nothing about the Internet changes the fundamentals of electronic commerce and its evolution. But it certainly enables every element of electronic commerce. The single belief underlying our guide is that electronic commerce is close to becoming the mainstream of business and a requirement for being a well-run business in the early 2000s. that will be so regardless of the Internet. But, the Internet opens up new opportunities and intranets are close to being a necessity for every firm.
|
