Digital Equipment Corporation: The Internet Company


Sirkka Jarvenpaa,
University of Texas at Austin and Marvin Bower Fellow, Harvard Business School
Blake Ives
Information Systems Research Center, C. T. Bauer College of Business, University of Houston

Table of Contents


Notice: Copyright, October of 1994, by Sirkka Jarvenpaa, Blake Ives,. The electronic version may be used by any not-for-profit educational institution. The case may be downloaded to university servers or for printed distribution but this statement and the authors names must remain attached and no changes may be made. The case may also be distributed in electronic form by the Digital Equipment Corporation (at no charge to users), again subject to this copyright statement remaining attached and no changes being made without permission of the authors. [Non-educational institutions wishing to download the electronic version to servers or to distribute the paper version must make arrangement with the authors.]

This case study is posted here at Samizdat Express with the permission of author Blake Ives.

Digital Equipment Corporation: 
The Internet Company (A)

"By the year 2000, every person in the developed world will have access to the Internet. Internet is kindred to the first personal computers. Another revolution is underway: internetworking all computers into a global information utility. The emergence of this 'information utility' is at least as important as the emergence of personal computers 15 years ago." - Sam Fuller, Vice President, Corporate Research


By the summer of 1994, Digital Equipment Corporation (Digital) had been a long-time user and supporter of the Internet. Digital's Internet successes already included: A number of individuals within Digital had pushed for the formation of an Internet business. These included people drawn from research, engineering, sales, corporate communications, consulting, and product and industry groups within Digital. Many had taken on "virtual" responsibilities above and beyond their normal jobs to help register Internet opportunities on the radar screens of senior executives. But many questions remained unanswered. How big are the business opportunities of the Internet and Web-based technologies? How should Digital compete in the Internet market? How to add value and make money from a resource perceived as free by many? How hard, fast, and deep should Digital invest in this market given the difficult financial state of the company? How should Digital further leverage the technology internally?

Digital Equipment Corporation (Digital)

Digital was founded in 1957 by Kenneth Olsen, a graduate in electrical engineering from M.I.T. Digital's inaugural product was the world's first inexpensive minicomputer, the PDP. Its successor, the VAX, established Digital as a computing powerhouse through much of the 1970s and 1980s. At its peak in 1989, Digital employed close to 130,000 employees and had worldwide revenues of $14 billion, over half of which came from outside the United States.

Besides instilling a strong engineering culture, Olsen put in place a matrix management structure. The matrix required that all business and industry groups (such as heathcare, media, personal computers, telecom, etc) build consensus with various functional groups - engineering, sales, marketing, and manufacturing - to launch products and services. The matrix allowed Digital to leverage scarce functional resources during the growth eras of the '70s and '80s.

The 1990s had downsized the computer industry; Digital was no exception. Between 1991 and 1994, the firm lost over $4 billion. At the end of the 1994 fiscal year (June), Digital reported yearly operating revenues of $13.5 billion, down 6% from 1993, an annual net loss of over $2 billion (including restructuring charges of $1.2 billion), and 85,000 employees. The firm's cost structure was higher than average for its industry. Gross margins on products had declined due to pricing and a continued shift in mix to low-end, low-margin products.

In the summer of 1994, Bob Palmer, the president and CEO, publicly committed to bring the firm to profitability by the end of 1994. To meet this goal, he implemented a simplified and less costly management structure. He also announced plans to reduce the headcount by another 20,000 by the end of the year and to sell noncore businesses.  The matrix management structure was replaced by one in which product divisions owned their functional resources to increase executive accountability, reduce high corporate operating costs, and speed decision making. An organization change also moved the systems integration business and information technology consulting practice within the computer systems division. As of July 1994, the head of the computer systems division planned to cut 14,000 employees from the division's 35,000 employees within the next 12 months.

During the summer of 1994, Digital published full-page ads in major business newspapers on its comeback in key businesses: "In the past year, we've seen our storage business grow 30%, our PC business grow 100%, and our Alpha AXP sales increase 164%."

In 1994, Digital's strengths were in client/server computing, networking technology, and systems integration. Digital's product line included scalable Alpha AXP systems, storage,networking, software and services, as well as industry-focused solutions that were increasingly offered in partnership with third parties. Computer systems sales accounted for 54% of revenues while services and others accounted for 47%. Although Digital continued to have a direct sales force targeting to the largest 1000 customers, approximately 45% of its product revenues in 1994 were generated through indirect channels, up sharply from 33% the previous year.

Digital and the Internet

Digital was a longtime user and supporter of the Internet, having established its first connection in the early 1980s (see Appendix A for background on the Internet and the WorldWide Web ). In the summer of 1994, Digital supported one of the most active gateways to the Internet network, handling over 2 million messages per month. The company had over 40,000 computers connected to the Internet and about 125 internal Web servers. Some servers were accessible to external users while others, so-called internal Web servers, were contained within Digital's electronic firewall and hence inaccessible to outsiders. The Internet with its open nonproprietary technology operated in marked contrast to Digital's past proprietary network technology, DECnet.

Digital also provided various support systems and services for Internet users. For instance, Digital's systems acted as relay points for newsgroup communications. Digital had also developed public domain software including FTPmail, which enabled users with limited electronic mail capabilities to access remote files. Digital was also a sponsoring member of the nonprofit CommerceNet consortium, which was building an open, Internet- based public information network. Among CommerceNet's objectives was the development of services and applications that met the reliability, security, and access requirements necessary for commercial exploitation of the Internet.

The Internet had initially been used primarily by research labs and engineering groups. Digital's first gateway to the Internet resided in the Western Research Lab in Palo Alto which had been a longtime supporter of Internet protocols and open standards computing. According to its director, Richard Swan, "Our interest has been with technologies that our colleagues in universities use rather than technologies developed by standards bodies that take 10-15 years to materialize."

Brian Reid, director of Digital's Network Systems Laboratory in Palo Alto, California, also valued the university connections that the Internet provided: "From universities we get the fresh talent and fresh ideas that help us continue to innovate. In the early days, corporate research was the only group within Digital connected to the net. We used to joke that we had more to talk about with the universities than with the rest of Digital."

The First Internet Product: S.E.A.L.

The first major stimulus for Digital's interest in turning the Internet into a business opportunity emerged from its long involvement with network security. An internally developed solution to the security problems turned into a potentially lucrative Internet business opportunity. The product offered a solution to one of the major barriers to commerce on the Internet: the perception (and reality) of the Internet's lack of security.

Brian Reid had been involved with network security long before joining Digital in 1985. As a member of Stanford University's faculty, he had been commissioned to help build the first firewall in 1983 so that Digital's newly formed research lab in Palo Alto could be connected safely to ARPAnet, a government-funded forerunner of the Internet.

Safeguarding Digital's network became a critical issue in the mid-1980s, when hackers initiated a series of attacks on Digital's network. A book, Cyberpunk, explained why Digital had become a desirable target of electronic espionage and sabotage: "Digital's Western Research Lab in Palo Alto ... was one of the few UNIX heavens.... High-speed network gateways may have been common at large universities and research laboratories, but they were less frequent at large corporations. Most IBM sites have no direct connections to the outside world. The Digital gateway ... with links throughout the world, was extraordinary in its very existence" (p. 117-118).

Digital had to either solve the security issue or disconnect itself from the network along with its direct links to valuable university contacts. Security became a major agenda item for Digital's network researchers. Network security was accomplished through technical innovations like gate-keeping computers as well as internal consultants and attention to manual procedures. Reid had himself written the first procedure manual in 1988. To him, the "firewall" was 15% mechanics and 85% procedure. An employee hired in 1988 developed advanced versions of the "firewall" using the concept of relay technology. One or more security computers stood between Digital's internal network and the external world of the Internet and constantly monitored incoming and outgoing traffic. Learning of Digital's secure interface to the Internet, customers like Dow Chemical had become interested in buying Digital's security system. Although not initially intended as a product, it had, on occasion, been sold as a service by Digital's consulting arm.

In May 1994, Digital launched the Screening External Access Link (S.E.A.L.) firewall, part of the comprehensive Internet Security Services. S.E.A.L. included a programmable gateway and security consulting service. It provided two-way protection. The intelligent gateway guarded internal private networks from any unauthorized inbound traffic. The gateway software could detect and audit access attempts from outside and limit outbound traffic. The gateway was able to filter data on a packet-by-packet basis, rejecting packets with a source or address that had not been pre-approved by an administrator. The gateway could also be programmed to provide controlled access inside the company to the Internet, or outbound traffic. For example, Digital's own firewall was programmed to allow two-way email and newsgroup communication, but only one-way (outbound) file transfer and WorldWide Web access and very limited one-way (outbound) Telnet usage.

The consulting part of S.E.A.L. included custom security consulting, security policy development, installation and configuration of customized security software, training in gateway management, and post-delivery telephone support. Cryptographic and authentication capabilities were optional. The base-level product cost $25,000.

Digital management expected S.E.A.L.-related revenues to exceed $50 million by the end of 1994. The target market for the product was chief information officers (CIOs), worried about connecting their private networks to the Internet. Reid painted a landscape for the demand: "There are tens of thousands of companies on the net. They all need a secure gateway linking them to the rest of the world. Many are going without, some are relying on home-made solutions, and others on mom and pop security services. If properly packaged and sold, the market for firewalls constitutes at least two thirds of the companies on the net."

Reid explained why it had taken so long to realize this business opportunity: "It took us over five years to convince management why they would want to sell a security firewall product. Selling network security was like selling dinosaur protection. Why do you need protection when there are no dinosaurs around? We had had no security violations for a number of years, and few people besides the research lab wanted to communicate outside of Digital. So why would someone want protection against a problem they didn't even have. The explosive growth of Mosaic and the WorldWide Web has, however, helped us make the case."

Swan emphasized the importance not only of selling the technical side, but also of managing perceptions about network security: "Many people do not recognize that they have a lower level of security in their businesses and homes than on the Internet. The S.E.A.L. gateway product is a security guard on the door; which most businesses do not have. Similarly, businesses doing business on the telephone have only a modicum of proof that the person they are talking to, or giving a credit card number to, is the person they think it is."

The WorldWide Web (Web) at Digital

Prior to the development of the Web relay through the firewall, Digital limited the use of the external Web (i.e., WorldWide Web ) to its own internal servers. Some employees, for example, maintained a "home page" inside the firewall which could be linked to from a previously unused field in Digital's online computerized employee locator facility (ELF). Personal home pages contained information about the person's background and interests as well as occasionally a personal or family portrait. The Cambridge lab was using the Web technology as a new nonpropriety, graphical user interface to old legacy information systems. No longer did a field engineer need to use specialized query languages to retrieve information from a problem-tracking and escalation database.

The first application of the web browser, Mosaic, at Digital was by Lucille Glassman, a technical writer assigned to Digital's Systems Research Center. In the spring of 1993, Digital was furnished with early copies of Mosaic. The initial application was a modest one, but it had far-reaching consequences. Glassman explained,

"Each summer college interns come to work for the three labs in Palo Alto, and we scramble to ensure that they can find documentation -- technical information about the programming environment and the current work going on in the lab, as well as non-technical information such as maps, train schedules, lists of restaurants, schedules for the bay area swimming pools, and so on. Someone in each lab had to pull such information together for their own interns - often compiling essentially the same information.

We brought up the first copy of Mosaic inside Digital a few weeks before the '93 summer interns arrived. Instead of updating our existing documentation, I chose to use Mosaic as the only information source for the interns. Very soon thereafter, Cambridge Research Lab and the Network Systems lab started Web efforts as well -- all of the Digital research labs were moving fast in the same direction. "

Besides providing worldwide access to technical reports, the hypertext links permitted related projects to be tied more closely together. For example, if a multimedia project in one lab depended on a research breakthrough in a storage project going on in another lab, the project plan for the multimedia project provided a Web link to the status report for the storage project. The Web also provided new ways to share research results. In the past, according to Glassman, "you tried to explain a 3-D animation project to someone by using words; now you link them directly to your application and let them experience it."

After the WorldWide Web relay was installed, and tested within Digital's Internet firewall, Digital permitted access to the external web (i.e., the WorldWide Web). Win Treese built and tested the Web relay. According to Jim Gettys, a senior scientist at the Cambridge lab, "It would have taken us a long time to get much synergy if we had to wait for these internal Web servers to provide critical mass. But once we opened up to the outside world, Digital people could see what was out there, what the possibilities were. Then it really took off. My colleague, Win Treese, used to say that it was like we found a snowball at the top of the hill and gave it a little push." In the research labs, Mosaic and the external Web were used to leverage expertise and tap outside resources to get things done internally. A web page developer within Digital explained, "Sometimes I just say to myself, I wish I had such and such a tool and then, voila, some college student somewhere - I call them my little elves - suddenly builds it and sticks it out on the web."

Lucille Glassman felt that the Web had transformed her job at Digital: "I used to be kind of stuck off in this funny corner of the lab. I focused on the people in the lab and usually on one aspect of a documentation effort - the beginner's manual, the technical report, or what have you. Now I communicate every day with people from all over the world -- and, now that Cambridge has solved the firewall problem, with people both outside and inside of Digital. And the projects I work on are much more integrated, with the boundaries between 'beginners' and 'experts' no longer important."

In another building in Palo Alto,California Russ Jones of Digital Corporate Communications was another early user of Mosaic. Part of his responsibilities included communicating Digital- related information over the Internet newsgroups and bulletin boards. Jones, like many others seeing Mosaic for the first time, reported, "After I saw Mosaic I rewrote my plans." He continued, "This is a technology that makes us look attractive by allowing us to project ourselves all over the world - directly to the screens of our customers." By October 1993, Jones had established Digital's external Web server. Customers around the world had access to over 3,000 documents, including technical manuals and research reports. The Web quickly became the place to get the most up-to-date information on Digital.

Digital was also the first computer industry vendor to offer online ordering of its full product line over the Internet. Customers could obtain descriptions of hardware, software, and services, read Digital's accouncements, configure systems, generate price quotations, and place orders. The service was initially offered to educational institutions and research laboratories, but by the summer of 1994, it was available to a broader community of U.S. customers, with plans for worldwide release. Digital also used the channel to deliver software patches. Potential customers could also test run their applications in the new Alpha AXP environment - a computer that was sold as a server for use with the Web and other internet applications. By early April 1994, more than 2,500 potential customers from 27 countries had logged onto these systems for a test drive.

It was one thing to get those already familiar with the Internet to embrace the new capabilities; it was another to sell it to Digital's management. Reid recalled, "I had been long convinced that Digital's proprietary networking architecture was dead. We either needed to get on board with open networking or get off the boat. However, that message was not getting through to headquarters. Then, one day, Lucille [Glassman] called me and said, 'Brian you have to see this!' I looked at Mosaic for two hours and knew I could now get management's attention. But I still needed to show them the profit potential."

Creating and Demonstrating Demand

Reid asked each member of his research team to go out and find a business to put on the Internet using the Web technology. One member, Glenn Trewitt was a big fan of science fiction and a good customer of the Future Fantasy Bookstore in Palo Alto, a supplier of science fiction and fantasy books as well as posters, pottery, and related paraphernalia. The store had long served mail-order customers, but that business changed dramatically when the store was put on the WorldWide Web. Jean Schroeter, the owner, or members of her staff, now spent several hours a day packing up orders for non-U.S. customers who had previously been out of reach of the store.

Communications about the store on the various Internet newsgroup's and bulletin boards provoked a flurry of interest in the store and an increasing stream of new sales. Another stimulus for sales was the mutual electronic links Schroeter had established with other Web businesses such as the Lysator Science Fiction and Fantasy Archive located in Linkoping, Sweden, which were likely to attract like-minded browsers. As of June 1994, 12-15% of Schroeter's sales came through the Internet. But the real payoff was to Digital. The Future Fantasy Bookstore was an eye-opener for many publishers.

Reid personally signed up the City of Palo Alto. Home to many of Silicon Valley's leading users of the Web, Palo Alto provided local visibility for Digital's efforts. And, as the home of Stanford University, a site for the '94 World Cup, it also provided international visibility. The Palo Alto home page included links to the city's services, pictures of parks, shopping the city council agenda, and library directories. Some of the local restaurants put their menus online, and a nightclub promoted its coming attractions. A local newspaper published an electronic version. The city's historical society provided a tour of how the city used to be. Future plans called for interactive services such as government forms distribution and bill payment.

The City of Palo Alto home page attracted so much attention worldwide that a Silicon Valley competitor sought to supplant Digital as the provider of the city's Web pages. Reid explained the pull-through from the City Pages; "Normally salespeople call customers. When customers call, they usually ask about fashionable things. Now, the second or third most common call we get in our sales offices is related to the customer interest spurred by the City of Palo Alto."

Richard Swan, the director of the Western Research Lab had similarly sponsored Internet-based projects. He explained, "Customers need help in visualizing the impact on their business. What amounts to a relatively small amount of time invested in tailored demonstrations goes a long way toward making a sale. I started a prototype newsletter to help people break out of the mode of passively looking at information on the Web. For example, one newpaper customer had a very elaborate description of how readers needed to go about submitting classified ads. Under the new scheme, the customers could do it themselves. This would lead to more ads and hopefully better ads."

Richard Seltzer, an East Coast member of the U.S. communications team, had also participated in a number of customer demonstrations. He recalled, "Representatives of a Dutch publisher were here the other day. After 15 minutes of a Mosaic demo, they started a discussion among themselves why their firm was investing in CD-ROM. We also had a large European consumer products company here. Before the end of that day, they were soliciting proposals from our salespeople."

Digital's Columbus, Ohio, office had worked with the North American Trade Point and the State of Ohio's department of Trade Development to create a virtual world trade center. Rather than build a physical export center, as other states had done, Web-based online electronic catalogs were built allowing Ohio-based firms to market their products around the world. The catalogs, running in a pilot mode in June 1994, allowed searching, ordering, and, in the case of a mail-order computer firm, customer-driven product configuration. Future plans included a trade-lead database.

Sandy Rupert, a member of the Digital team on the project, explained, "The Ohio Department of Trade Development, working with the North American Trade Point, which is also located in Columbus, Ohio, were looking for an electronic platform for their trade center. It needed to be easy to use by nontechnical people, useable on any platform, accessible affordably by nearly anyone anywhere in the world, and supporting of multimedia." Jamie Sisto, deputy director at the Ohio Department of Commerce, described the response from the participating Ohio companies: "Their eyes light up when they see their products being advertised all over the world." Sisto's organization took responsibility for the production of the catalog; but the maintenance of each firm's Web pages was the responsibility of each firm."

Meeting Demand for the Internet

The new business, however, faced fulfillment and delivery challenges within Digital. Aside from security products, there were no standard product codes for such things as "install a web server." Moreover, these new business opportunities did not match up with the large- projects business of Digital consulting. One observer noted, "We are at a disadvantage compared with our $200-an-hour consulting business. We are now operating in an environment where the U.S. president advertises over the Internet for an intern to come and develop Web pages for the White House."

And yet, more people inside and outside of Digital were excited about the Internet, and more and more people asked the questions of 'what can we sell?' Reid explained the supply-side problem, "We have a chicken and an egg problem. Salespeople are beginning to bring customers into the lab or to take me to customer locations. I have probably spoken with 45 CIOs in the past few months. We have had banks here and book publishers. They show up with cadres of people and lots of interest. So we have begun to create customer demand, but the question now is who is going to supply the solutions."

Till August 1994, Digital consulting served as the primary formal arm for delivering Internet related business. In reality, however, Internet solutions were often being provided by researchers and engineers familiar with the Internet who were working side by side with sales support personnel. In many instances, they did so despite their formal job assignments. The Web had begun to change the role of corporate research, as Sam Fuller explained in June 1994: "In the past when research came up with an idea, they passed it on to a product development group that carried the idea forward. It is less true in this case. Because the Internet reaches directly into the marketplace, Corporate Research has been able to quickly prototype and deliver Internet solutions within days of the initial customer contact." One manager at Digital noted, "our R&D people are turning into product entrepreneurs."

Swan also remarked on the new relationships that the Web technology had created, "This is a technology that leaps very quickly from creation to use. Direct relationships are developing between salespeople and researchers, that often bypass the traditional engineering and product release organizations. We are likely to see more of that in the future as the value to the customer moves increasingly away from base technologies - chips, iron, operating systems - to applications. This is akin to small companies where there is inevitably a close relationship between engineers, designers, and customers. In general, researchers seem to enjoy this close interaction with customers."

Interactions with customers also changed the problems that researchers wanted to work on. According to Swan, "Some of my people are working with a number of cities that want to put hazardous materials online. Just as you might use the Web to cruise down an electronic street in search of a restaurant, a fireman could cruise down a street on the web to find out what hazardous materials were stored there. The lab is now building a cellular paging version of this so that a fireman can use Apple's Newton handheld computer to access pages of the Web."

With Internet technology, customer applications could be built at costs an order of magnitude smaller than in the past. For example, an application was being built that helped track criminals moving from one city to another in a ten-city area. According to Swan, "You could have built such a database in the past, but the cities could not have afforded the infrastructure to implement it. A proprietary network would have cost them millions. Now a police station can get access to a Web for $25 a month, or $1,000 a month if they want more bandwidth. So, now the problem is building the database rather than investing in the infrastructure."

Selling Digital Management

Since October 1993 Digital had used the Internet as a virtual site, along with Boston, Paris, San Jose, and Tokyo for new product announcements. Product info was made available on the Internet and pointers to that information in Digital's home page and inadvertisements placed in the Global Network Navigator, an emerging electronic magazine on the Web.

Sam Fuller directly attributed $100 million in 1994 sales to Digital's product information and benchmarking services available on the Net. The challenge, however, remained to push the corporation to recognize the growing business opportunity and to act upon it quickly. Digital's precarious finances necessitated a business case based on few dedicated resources. And approval of the plan required that management view this as a solution to the crisis the corporation faced.

Fuller had first brought the Internet to senior management's attention in early 1993. After taking over as CEO, Bob Palmer wanted an update on the key trends in computing. In his presentation, Fuller showcased the Internet: "I told them that the Internet was the most important thing happening in computing. In my talk I took them around the Internet showing them the resources available. I tried to help them understand what this might mean to us by the year 2000. When they later saw Mosaic in the fall of 1993, they could envision specific products and services we could sell by 1994 and 1995. And it didn't hurt that in 1993, they were reading something about the information highway in almost every newspaper or magazine they picked up."

Reid had spent much of his time since the summer of 1993 pushing the Internet, the Web, and Mosaic. Both in the fall of 1993 and the winter of 1994, he and his lab members showcased their Internet-related work to engineering and research staffs on the East Coast. The fall meeting had raised awareness, while the winter meeting demonstrated to the East Coast staff that unless Digital soon acted on the Internet opportunity, there would be plenty of competitors anxious to enrich their own corporate coffers. Reid's staff was accompanied to Massachusetts by Sheridan Forbes, a sales representative from the West Coast. Her formal role was the Call Center Plus product line, but she was spending more and more of her time selling Internet-related products including Web consulting, Web solutions, and secure mail gateways. Her target industry was entertainment where the Web technology could be used to dramatically reduce cycle time, particularly for projects involving subcontractors. At the meeting Forbes provided impressive estimates of the revenue potential of Internet business in her target industry.

Reid and Fuller had personally demonstrated the Web to many senior Digital executives. Reid noted, " There are 175 VPs in this company; I have brought the Internet message personally to half of them in face-to-face meetings. Fuller has personally brought the message to senior VPs and business unit heads. With the Internet technology there is no substitute for hands-on experience." Swan agreed, "People who have never used the technology can't imagine what the fuss is about."

But one-on-one demonstrations were a slow way to educate an entire company. Mass conversions required mass media. A video, initially prepared for a sales rally, became the cornerstone for promoting awareness broadly inside and outside of Digital. This three-minute video, prepared by Richard Seltzer, toured some 100 home pages and reflected Digital's position at the forefront of Web-based business transformation. Seltzer explained: "The video demonstrates that the Internet is much more than electronic mail, newsgroups, and bulletin boards. Global electronic commerce is happening on the Internet. A lot of players are congregating. While communication vendors prepare us for a day in the future when we might have this or that, the tape shows it is already here, and Digital, a leader, is already there. This is a new way to do business." The sales force worldwide saw the video, as did numerous customer groups, market analysts, and even competitors. [You can now see and hear this video, "A Glimpse of the Future" online  broadband (256K), dialup (56K); requires RealPlayer. You can read the script at http://www.seltzerbooks.com/glimpse.html.]

Getting beyond awareness was another challenge in the fast moving marketplace. Information given in a previous debriefing might be totally invalid by the next debriefing. Some also did not appreciate the evolving nature of the Internet.

Executive support began to mount in March 1994. Bob Palmer, in his keynote presentation at the UNIX trade show, UniForm, showed the videotape and did an online demonstration of Mosaic and the Worldwide Web using a laptop. The presentation was attended by several thousand people including a number of CEOs. It generated a flurry of sales leads.

According to Russ Jones of Corporate Communications, "Innovations like this do not originate from the top of the organization. The CIO doesn't say, 'Hey, I have been thinking about this Internet thing' - though that is finally beginning to happen. Rather, the employees out on the frontline realize the value of the new innovation and keep pushing it. One has to keep building a critical mass of passionate believers and then keep moving higher and higher in the organization to get management buy-in."

* * * * * *

Building the Business Case with Virtual Teams

Establishing a business around the Internet required lots of human resources. But, as one activist stated, "None of us had anything on the Internet written in our job descriptions." Given the financial state of the company, timing also became critical. Jim Neumann, the focal point of Internet business within Digital consulting, explained, "We struggled over when we should reveal, and then start pushing, the business plan. Everything has its time and circumstance. The company did not have a lot of resources. If something was presented and got rejected, it was going to be hard to revitalize it later on."

Palmer's demonstration of Mosaic and the WorldWide Web convinced many that the time had come to establish an Internet business. Virtual teams, dealing with both strategic and tactical issues, started to meet regularly in mid-March. Members of virtual teams took on responsibilities other than their formal ones. Digital's old matrix structure had nurtured a culture in which virtual teams were common. In the past, people could turn their virtual job into a real job. Now, however, facing serious financial and market challenges, it became riskier to devote time and energy to something that was not established formally. Those who contributed might lose the support of their regular boss and, by so doing, become candidates for layoffs.

Under Fuller's leadership, representatives from new market and business development, engineering, publishing, consulting, research, and so on came together to focus on developing the business case. Another virtual team focused on its delivery. The latter team had overlapping membership with the first one, but it also included individuals from communications, the NT windows organization, network engineering, educational marketing, security services, ChannelWorks marketing, etc. Since the team members were geographically distributed all across the count ry, the virtual teams used communication technology (electronic mail, teleconferencing) as well as face-to-face meetings to communicate.

Ellen Kokos, VP of market development of Digital's Advanced Technology Group, was an active member of the virtual teams. Her emerging technologies group had done earlier a study on what Digital ought to do with interactive media and had recommended attacking the Internet market. According to Kokos, "Sam Fuller put together a dynamic study group - the informal Internet Business Working Group - to determine what it meant to attack the Internet market." By the end of April, the group developed a business case and a strategic plan that outlined (1) the products, (2) a marketing plan, and (3) an organization that would support the business.

The new organization's responsibility was to leverage Internet expertise within Digital. According to Kokos, "Up to this point, the Internet business community was an informal one where everyone marched to their own drum. To be successful, there needed to be a common strategy and an implementation plan." Another participant had a different view of the importance of the Internet organization: "There is a limit to what you can do with a virtual team. We have been swarming after the ball with great enthusiasm, but we need to start scoring. An Internet business manager is needed to get dedicated resources and to legitimize us."

Building Toward the Internet Business Group

The Internet was not a clean fit with any of Digital's existing business units, although it would have implications for most. The Internet business could, according to Reid, have been structured as another "stovepipe functional business" or it could be established to run horizontally across the existing businesses - thus being more a reason to buy other products (consulting services, servers, workstations) than a marketplace by itself. Reid favored the horizontal structure for he felt it would spur demand for other products and avoid interdivisional competition.

Some felt that Digital was at somewhat of a disadvantage with competitors like Silicon Graphics and Sun whose internal management information systems were based on TCP/IP technology. One Digital executive explained, "When a customer calls their office, even the secretaries can talk about Internet and Web-based solutions because they use it every day in their own activities." Although nearly all of Digital's employees had access to email and newsgroups on the Internet, much of electronic employee communications were still via the proprietary DECnet communications architecture. As a result, many customer inquiries related to the Internet and the WWW ended up in Digital's research centers.

At the end of May, the consulting division conducted a worldwide semiformal Internet skills survey. Jim Neumann, a member of the Internet Business Working Group from the consulting division, described the results, " We learned that a lot of the employees had gotten skills because it was of personal interest for them to do so. We found more Internet skills than we expected."

Besides struggles over the shape and skills of the possible organization supporting the Internet, questions began to arise regarding the viability of the business model. Neumann, a member of the Internet Business Working Group, recalled, "Our initial business model assumed that people would need our services to access the Internet, to protect themselves, to redesign networks, and to set up Web servers. But, this was only the tip of the iceberg. Web servers brought customers to the table, but the real business opportunity was revealed by the questions that followed. Those questions dealt with business reengineering, education, and integration with legacy systems."

Neumann continued, "We also established that businesses wanted to do this in bite-sized chunks. Internet technology allows customers to start small and build up slowly. To get started, customers don't need to buy additional machines. Those are only needed if one scales up. The $250,000 deals in this business appear in the sequence of $25,000. Besides the chunked revenue stream, some areas in this business have very low entry barriers. Already there are some 20 to 30 businesses offering web service outsourcing."

By the end of August of 1994, the estimated number of Internet service providers was in excess of 250.

Ellen Kokos had a related view: "The real Internet opportunities are with those who want to buy and sell information, products, and services like you and me at our desktops and in our homes. What is valued will be making all that easy to do. The money to be made here is in high volume commodity products. But it is also a challenging mass market because it is diverse, nonhomogenous, and involves multiple operating systems and platforms."

Kokos also outlined the landscape of the Internet-related business: "There are three legs to this business. The first is connectivity, or providing the store and forward packet capability. AT&T or MCI in partnership with others have a real opportunity to dominate here.The second leg of competition is the content providers. These range from Dun and Bradstreet, the Wall Street Journal, and you and me. Everyone will want to be a publisher. The third leg consists of the commerce enablers who have expertise in transaction management and processing. Digital can help get people ready for the Internet, particularly the content providers and commerce enablers. We can also provide security. We now have S.E.A.L. for businesses. We can come up with home S.E.A.L."

Jim Gettys, at the Cambridge Research Lab, felt that Digital was three to nine months ahead of other large computer companies in the Internet marketplace. However, staying ahead of the pack required an ability to innovate at an unprecedented pace. According to Gettys, "As the New Yorker cartoon said, 'Nobody knows you are a dog on the Internet.' But if you are going to displace paper catalogs with electronic ones, they had better provide things paper cannot. Customization will be a big piece of this. Retailers will bring your attention to things you want, for instance, items like those you have bought before, rather than try to sell you things that you probably don't want or need.This will also help retailers manage their inventory. If a new catalog item isn't selling after the first day, it probably isn't going to sell well so don't buy more. If you are about to sell out, take it off the catalog. Imagine the power of Consumer Reports if there was a hypertext link directly from their best product designation and that firm's electronic catalog."

There were also many challenges Digital had to overcome. It was also unclear whom to sell the various services to. Richard Swan explained, "Our salespeople find that they are selling to the top of the organization; MIS people are very much off to the side. Newmann had similar experiences: "CIO types see this as interesting, but outside their domain of expertise and responsibility. Business people are much more likely to start to rock and roll after our presentation. They also have a much easier time conceptualizing what this means to their companies."

Internet product design also continued to be a challenge. Kokos explained, "The Web technology becomes increasingly exciting when you add some interactive media capability. However, you still have to design your products to reach the widest possible audience.We need to remember that everyone does not have a 486 on her or his desk at home or in the office."

There was also still a lot of customer concern about how secure and reliable the Internet was, who controlled it, what the real costs were, and how, or if, to advertise? Many manufacturing organizations, in particular, perceived that since their customers were not pushing the frontiers of Internet technology, it was safe to sit still.

The new business processes that would allow firms to take advantage of the Internet and Internet related technologies were still very much a black box. Those willing to experiment had to be also willing to make the investment to learn how the Internet was different from other delivery channels. A member of market development explained, "A pull model, rather than a traditional push model, is likely to dominate. Companies have to become smarter about how to package their services in a series of hierarchical packages. The Economist magazine provides a useful example. Their message is, 'You can get a subscription for a mere $100 a year. But you really want to get the business intelligence unit, and that is just another $300 a year. Also, here is a list of focus group reports that you can subsribe to for $1,000 a month. One has to advertise by pushing free teasers out and let customers pull for the revenue-generating items."

The challenges were accompanied by high hopes. Richard Swan reflected on the promises: "the Internet has gotten a lot of intense publicity, but nothing can get this much publicity forever. The pendulum will swing back. But the difference between the Internet technology and other technological wonders is that this technology is here today and is not going to go away. It is as fundamental as fax machines, phones, airlines, and trains."

Not only did the Internet mean a convergence of media, it also meant a convergence of industries. Sam Fuller explained, "The boundaries that once separated information industries like computing, communications, and consumer electronics are disappearing.... The result will be revolutionary changes in the way we deliver, re ceive, store, manipulate, share, and respond to information."


Internet, Worldwide Web, and Mosaic

Since the late 1960s, Internet and its predecessor ARPAnet, linked scientists and academics around the world. By 1994, the Internet had become a business network as well as an academic and research network. The Internet, a network of networks, was a set of interoperable communication protocols and gateways (now called routers) combining over 3.2 million computers around the world. By the middle of 1994, over 50% of Internet traffic was from commercial firms and on a profit-making basis. The number of non-US networks exceeded the number of US networks. Internet ran over any communications medium: telephone twisted pair, cable, fiber, or cellular wireless.

In 1983, the U.S. Defense Department required that any user of ARPAnet use the Transmission Control Protocol/Internet Protocol (TCP/IP), a public-domain protocol designed for peer-to-peer communications among network nodes. In the TCP/IP environment, clients rather than servers (as in traditional hierarchical communication architectures) controlled the interaction. Among the functions supported were abilities to execute a procedure running on a remote computer (remote procedure call), to access files located elsewhere on the network (FTP or file transfer protocol), to find people and functions located on other nodes (directory), and to authorize access to functions and data (implemented through a security server). In 1994, over 70 countries were assumed to have full Internet connectivity with another 80 with limited forms of connectivity. Increasingly, commercial off-the-shelf software products included TCP/IP compatibility.

In the summer of 1994, the WorldWide Web (Web) was the fastest-growing Internet information service.  The Web was a diverse network of server computers providing a limitless distributed information service based upon the hypertext model of information representation. Information (text, graphics, video, images, and/or audio) was stored in documents that had hypertext links to other documents on other computers. The Web was initially developed by physicists and engineers at the European Laboratory for Particle Physics (CERN)in Geneva, Switzerland, to pool information and knowledge from individuals in various remote sites who collaborated on common research projects. The Web was designed to allow scaling and ecological growth of information. Each user and project could prepare their information independently, and as relationships emerged, those could be reflected without massive centralized changes to the existing information structures.  The Web accommodated the automatic construction of menus and directories. It was also designed to be expandable and upgradable, accommodating new multimedia formats as they were invented. Technically, the Web could be defined as an (1) addressing system , (2) a network protocol, and (3) a hypertext markup language.  Anyone wanting to be part of the Web had to use the three standards.

In 1994, commercial firms used the Web to house technical data of value to their employees, business partners, and customers. Many had also established an electronic store front that provided sales and marketing information accessible by customers anywhere in the world. Online ordering was sometimes also available. For example, one clothing store prompts for one's size, and points to a virtual store containing only those clothes that are the right size and also in stock. Such store fronts faced no geographical boundaries.

Although the Web had existed since the late-1980s, it remained a sleeping beauty until the arrival of Mosaic, developed by the National Center for Supercomputer Applications (NCSA) at the University of Illinois at Urbana-Champaign. Mosaic, a highly user-friendly graphical point and click software "browser" that resided on the user's workstation, was often called the "killer application" of the Web because it made it accessible to nontechnical people. Early copies of Mosaic were distributed in early 1993 to universities and research labs. By the end of 1993, Mosaic was becoming a household word in those communities. By the summer of 1994, the business community recognized the name, though it remained largely a myth to those with no firsthand experience.

Mosaic could access, retrieve, display, store, and forward documents and information from any authorized Web server in the world, delivering it in the same familiar format each time to the user. It did so regardless of the underlying computing hardware or operating system technology of either the web "server" (where the information to be accessed resided) or the client workstation. No longer did the user need to master different command languages to retrieve information from different networks or databases where Web documents resided. Bob Palmer, president and CEO of Digital, summarized the technology as follows: "Mosaic and the WorldWide Web - a distributed, user-driven application that roams the Internet and draws on a set of databases from all over the world - are the beginnings of an information superhighway, if you will." By the summer of 1994, Mosaic had been enhanced to facilitate electronic commerce (i.e., forms capability, encryption, graphical mapping, and signature authentication).

To many, the Internet and the WorldWide Web were fast becoming an early working model of an information superhighway.  Others, however, believed that the superhighway would emerge through cable and interactive TV technology. Richard Seltzer, an early proponent of the Internet at Digital, explained the difference he saw between interactive TV versus an Internet-based information highway: "Interactive TV would have been perfect for Hitler. You would have everybody wired and give them limited choices. To excite them as a mass mob, show them some massacre in Somalia and give them 30 seconds to decide if we should nuke them. There is no discussion or limited discussion. The Internet is the opposite in that everybody can say everything about everything. Everyone's view is heard. It is a lot more democratic. For the most part, names don't matter. Ideas do." Another executive at Digital recalled how one customer in a presentation had jumped up and called this the great equalizer between developed and developing countries. From there on, the discussion took on the flavor of social policy. Russ Jones, an Internet program manager of corporate communications held this view: "There appears to be a lot of convergence between the national information highway and the Internet. When people talk about what the benefits are - the virtual shopping malls, the ability to browse through catalogs and get at and share information Ð about 80% of those benefits can already be derived from the Internet , putting aside performance constraints that some users suffer."

seltzer@seltzerbooks.com privacy statement