Chapter 5 - Data Communications: Market Competition 1969-1972
5.1 Entrepreneurism Flourishes 1968-1972
By the end of 1968, the economic opportunities in data communications from the increasing use of computers by corporations, the certainty of coming market competition in telecommunications, and the unprecedented capital available to be invested in high technology companies, all combined to fuel entrepreneurial ambitions. From a mere handful of competitors in 1968, over one hundred firms would be selling modems and multiplexers by 1972. The following firms would join Codex, Milgo and ADS in eventually seizing market leadership and market share from AT&T.
In 1967, a frustrated Stan Hunkins challenged his boss, the vice president of engineering of Ultronic Systems. Hunkins said he could assemble a volunteer team and design a stock-quote desk system in two months, a task the company had not been able to accomplish in over two years. Given a green light, they succeeded. Afterwards the team members hungered for more such opportunities and freedom. However, the acquisition of Ultronic by General Telephone & Electronics (GTE), a $2 billion company, portended more bureaucracy. The conversations of the group soon turned to forming their own company. In September 1968, Hunkins, Jim Hahn, Joe Andrews and Tony Barbaro launched Infotron in Hunkins’ basement. From their Ultronic background, they knew that companies wanted to interface large number of asynchronous terminals to computers. So they decided to innovate a time division multiplexer (TDM). In 1969 they introduced their TL110, so named because it multiplexed teletype machines that ran at 110 bits per second (bps).
Like the founders of Infotron, Charles P. Johnson did not need others to convince him of the opportunities in data communications. Working early in his career at Illinois Bell Telephone, then selling telephone services to the SAGE project for International Telephone and Telegraph, and finally working for the military contractor Stelma, Johnson understood the need for modems and multiplexers. After Data Products Inc. acquired Stelma in 1968, Johnson decided it was time to strike out on his own. Needing a team, he approached others dissatisfied with the Data Products acquisition. Over the winter of 1968, Johnson, Robert Smith, Tom Lehrman and Jack Arcara wrote a business plan. Smith remembers:
At that time, in the commercial world, you basically had your telephone network and some private line teletype nets that were supplied by Western Union or AT&T. The way you fixed things is that when something went wrong, you picked up the phone and called Western Union or AT&T and said: “My circuit’s down. Fix it,” and you waited for it to get fixed. For administrative teletype nets, low speed stuff, that was probably ok, but now you’ve got this huge computer center, and there’s going to be remote terminals – timesharing was beginning to come into vogue by that time – and that was not enough. We wanted to provide modems and multiplexers that not only could pass data through, link to link, but also have some diagnostic capability. We wanted to provide the user with some information as to what was happening. So there were some philosophies that came out of our background in the military systems and these big networks that were the foundation of the business plan.
In 1969, the Johnson-led team formed General DataComm Industries, Inc. Incrementally innovating on the widely successful ADS-660 time-division multiplexer (TDM), they introduced their TDM at the International Communications Association trade show in 1970. They would have sales of $160, 000 in 1970 with a loss of $676,000. Western Union would be their first customer, and remained their most important customer for many years.
In 1969, Ed Botwinick, an analyst with the investment bank Goldman Sachs, came across a business plan written by a group of engineers wanting to leave Western Union to start a company selling TDMs (Sidney Kaplan, Richard Schmal John Elich). Having passed on making an investment in ADS a year earlier, Botwinick had maintained an interest in TDMs and met with the engineers. He remembers:
Their business plan didn’t really make sense, but their product plans did, so I became a member of the founding group, and I rewrote the business plan and raised all of the start-up.
With an initial investment of $210, 000, Timeplex began operations in 1969. In 1970, struggling to make sales, Timeplex signed a take or pay agreement with industry-leader Milgo. They ended 1970 with sales of $109, 000 and a loss of $205, 000.
In 1969, a group of engineers led by Joe Looney left Control Data Corporation and founded Paradyne. Backed by venture capitalists, they pursued a product strategy of building modems and front-end data communication processors to be connected to mainframe computers. For five years they would struggle before becoming a leading modem manufacturer.
In early 1969, Kim Maxwell and six others founded Vadic, one of dozens of companies that would jump into the dial-up modem business. At first, Vadic planned to develop both control systems and modems because they had one person who knew a little about each. But soon they realized they needed more focus and collectively decided it would be modems. Maxwell remembers:
We looked around and then Carterphone sort of began to dawn on us as we went and talked to some potential customers. “What kind of modems would you like?” Well, people said things like Bell 103s. So we decided we better go find out what a 103 was. We bought the technical book from AT&T and we read through it. Now that tells you the level of knowledge and sophistication of this company that created the dial-up modem business.
Within a few years, Vadic would challenge AT&T as the leading innovator of dial-up modems.
Universal Data Systems
In 1969, Mark Smith, an engineer with SEI Systems, a government contractor, realized he did not want to work in an engineering job shop forever: he wanted to create products to sell. As he searched for what to do, he began reading about Carterphone. Knowing nothing about modems, he contacted AT&T to speak to a modem salesman. Smith remembers:
When the, quote, “salesman” came out and explained to me the reason that AT&T’s unit was more expensive and five times as big as the other units that were starting to show up on the market was that it had more resistors and capacitors in it, I figured that this might be a good business opportunity.
Later that year, Smith and a friend, John Howell, decided to start their own company, Universal Data Systems. Having but $30,000 between them, they kept their jobs and worked nights and weekends to build their first product, a Bell 202 lease-line modem. Smith remembers:
We did the Bell 202 modem simply because it was conceptually, in every way, the simplest possible thing you could do, and we wanted to get started with something. By August of ‘70, we had ten units that were basically complete. So we had our show and tell units. The first of August I resigned. We still had 25 of our $30,000 left by that time.
Hustling whatever opportunities they could uncover, they soon bid on a large job for the Southern Company, a major utility in the southeast. In October they were notified they had lost the bid to ADS. Nearly out of cash and an end to their entrepreneurial aspirations, they received an unexpected phone call in December. Smith recalls:
The program manager in Birmingham had gotten mad, for whatever reason, with ADS. He asked us if we’d like to supply the modems. We said: “Heavens, yes.” So we got a contract for maybe $200,000, and, in addition, we got a $50,000 advance. That put us in business.
The surge in entrepreneurship was not limited to data communications. New firm formation also reshaped the nature of competition in software and minicomputers.
In the 1960’s, thousands of new software firms were founded, mostly after 1966.1 The real stampede occurred in 1969, after IBM’s announcement in December 1968 implying that they were going to unbundle software, which they formally announced in June 1969.2 While many speculated that the antitrust lawsuit the Government filed against IBM on January 17, 1969 prompted IBM’s decision, the more probable reason was software was becoming prohibitively expensive to develop and maintain. Software sold in generic packages to many customers had not yet evolved. Instead, software had to be created from scratch for every application and was generally a massive undertaking, a view reinforced by the expensive and challenging development of the software for IBM’s System/360. The well-publicized IBM experience reinforced that of the first true software project, the SAGE project, which at its peak, in 1959, employed more than 700 programmers and 1400 support people, half the total software manpower in the United States.3 Yet after IBM’s unbundling announcement, entrepreneurs were ready to transform the opportunities in software into successful companies. Total sales of software, estimated to be only $20 million to $50 million in 1969, would soar to $400 million in 1975.4
Successful software companies invariably used data communication products as did the products they sold to users.
Also driving the need for data communication products was be the new category of computers called minicomputers. With a history that also dates to the SAGE Project, and is considered to formally have begun with the introduction of the PDP-8 by Digital Equipment Corporation (DEC) in 1965, an industry with only a handful of firms in 1968 exploded to more than one hundred firms with the entrance of ninety-two competitors between 1968 and 1972, sixty percent being new start-ups.5 Sales would skyrocket from $200 million in 1968 to nearly $1.5 billion in 1975. More than four times as many minicomputers would be sold in 1975 than mainframe computers, tens of thousands of computers potentially needing modems and multiplexers, especially after time sharing-based minicomputers started becoming available after 1976.
The transforming changes taking place in information technologies were staggering because of new standards in software; advances in semiconductor technology such as Large Scale Integration (LSI) technology, integrated circuits, and semiconductor memory; on-line, real-time computing such as timesharing; radical reductions in costs due to volume production; capital available both privately and publicly; and quickening use of computers by corporations, all leading to ferment and growth, albeit no one at the time had a crystal ball as to how it would all shake out.6 Adding to the confusion were the gyrations in economic policy meant to bring stability but resulting in a roller coaster lasting longer than anyone would have imagined, or certainly wanted.