Jon Peddie felt vaguely uneasy, perhaps even a little irritated as he sped along California's Highway 17 last October, on his way from Berkeley to Monterey. With him were two of his top managers from Jupiter Systems, a fledgling maker of high-performance graphics equipment, of which Peddie was founder and president. Their destination was a resort hotel overlooking the Pacific Ocean, where they were to attend some sort of business conference -- for what purpose Peddie was not exactly sure. He had made the commitment to go several weeks before, but now he had visions of spending two days in boring meetings followed by boring banquets while, back at the office, a major bid sat waiting to be completed. Quite frankly, the prospect made him apprehensive.

What was this meeting all about, anyway? Peddie knew little about the other participants, or their companies, or the businesses they were engaged in. What were they going to talk about? What did they have to say to each other? Why were they even being brought together?

Indeed, they were a disparate group -- the founders and managers of 10 young technology companies working in fields that seemed only tangentially related. Their products spanned the diverse frontiers of information processing -- from super-fast integrated circuits to specialized microcomputers to medical imaging systems. Some of the companies had annual revenues approaching the $20-million mark; others had no revenues at all. But appearances notwithstanding, the companies did have at least one thing in common: Each had recently sold a minority equity interest to the host of the two-day gathering in Monterey, a $214-million, Massachusetts-based electronics company by the name of Analog Devices Inc. And therein lies a tale.

Founded in 1965, Analog Devices was in the business of making electronic components for scientific and engineering applications. Over the years, it had managed its growth well, seizing diversification opportunities as they presented themselves and developing most of its new products from within. But about four years ago, as the pace of technological innovation began to quicken, Analog founder and chairman Ray Stata looked ahead and saw trouble. The market for the company's electronic components was changing, and changing fast. Within a few years, new products would be available with capabilities that would far surpass those of Analog's old products. Unless the company found better ways to respond to this challenge, he feared that the days of impressive growth would soon be over. "To do what we needed to do," Stata explains, "we had to regain some of the facility and flexibility of a much smaller company." The question was, how to do it?

Stata didn't believe that Analog had the internal resources to respond effectively, and he knew that he could go out and acquire a bunch of young companies only at the risk of losing key talent through defection. Instead, he decided to try something different. "Our idea was to create a series of relationships with promising companies," Stata says.

By "relationships," Stata meant quasi-partnerships -- arrangements that would serve both Analog and the companies with which it became involved. Each relationship would have to be nurtured individually, in keeping with the interests of the small company founders and their key employees. Some of the companies might eventually be acquired; others would not. If the approach was going to work, notes Stata, "everything had to be built on mutual benefits."

All of this could, of course, prove expensive. Somehow Analog would have to find the cash to finance the outside investments without undermining its own ability to fund internal product development. Stata and senior vice-president Larry Sullivan decided to approach Standard Oil Co. (Indiana), which was a minority shareholder in Analog, and one with diversification ideas of its own. Standard Oil agreed to advance Analog $10 million a year to invest in promising young technology companies. In exchange, the giant oil company would receive options to buy additional shares in Analog at a future date.

With Standard Oil's money, Analog established relationships with the 10 companies that gathered in Monterey last October. In a sense, the companies represented Analog's vision of the future, a vision that was still in the process of being defined. With that in mind, Analog had called the conference -- to give the far-flung entrepreneurs a chance to explore the ways in which they might learn from each other and the directions in which the various reIationships might lead.

The entrepreneurs soon realized that they were part of an unusual experiment. Meeting in a second-floor conference room of the hotel, they began to feel a certain camaraderie. "It was almost as if we were members of a club," Peddie recalls. But what kind of a club? Were they simply pieces in an unorthodox venture capital portfolio? Were they future candidates for acquisition? Or were they pioneers developing an entirely new type of corporate entity? Nobody seemed to know.

Indeed, everyone shared that sense of exploring the unknown -- even Ray Stata, who stood up to make some after-dinner remarks following the first day of meetings. "Many of us expected him to raise the flag and talk about how Analog reached the $200-million-mark and grew to 3,000 employees," says Peddie. Instead, Stata presented himself as another entrepreneur who, like everyone else in the room, was still searching for solutions. "He told us that he was still muddling through problems one step at a time, much the way he did when he ran a $5-million or $10-million company. And he said something that made an impression on a lot of us. To be successful, he said, you couldn't just depend on doing things better. You also had to be willing to do things differently."

Sitting in his office in Norwood, Mass., 20 miles south of Boston, Stata elaborated on this message: "I wanted to get across the sense that even as a company gets larger and more mature, it never really gets there. You become more professionally managed, only to find that you need to become more entrepreneurial and less formal. There's always another challenge around the corner."

For Analog, the immediate challenge was to learn how to work with a group of developing companies, harnessing their entrepreneurial energy without undermining or crushing the creative forces within them. It was a task that had stymied many large corporations in the past, as Stata was well aware. Analog had one distinct advantage, however: It had been a small, entrepreneurial company itself not so very long ago.

In many ways, the Analog experiment had its roots in the early history of the company -- going back to 1965, when Ray Stata teamed up with Matthew Lorber, his former roommate at Massachusetts Institute of Technology, to get into the electronics business. Their idea was to launch the company by producing components known as operational amplifiers (op-amps) -- devices used by military contractors, medical-equipment makers, research laboratories, and others to process signals that measure real world conditions, such as temperature and pressure, with a high degree of accuracy.

At the time, preassembled op-amps were in short supply. People who needed such components often wound up making their own. Stata and Lorber figured that they could use their off-the-shelf models as a wedge into the market, allowing them to form strong relationships with customers. "We saw an opportunity to create a very broad customer franchise and to build our name in a particular corner of technology," says Stata.

Given that strong market orientation, it was essential for Analog to have a highly skilled sales force, and Stata proceeded to hire trained engineers to call on prospective customers. Not only could these engineers explain the precise performance parameters of Analog's existing components, they were also knowledgeable enough to alert Analog to the changing needs of customers and to suggest new components for the company's product line.

Back in 1967 and '68, for example, John Harwood, one of Analog's first sales representatives, had a customer called Technicon Instruments Corp., in Tarrytown, N.Y., a maker of electronic medical equipment used in the clinical analysis of blood. In the course of making his sales calls, Harwood became aware of a problem Technicon's engineers were having in the design of an upcoming product. What they needed was a new component that didn't exist -- one that could convert readings of light into numbers. Harwood relayed the information to Stata. "I told him about my customer's projected requirements," says Harwood, who still sells for Analog in the New York area. "It was Ray who decided to commit the resources."

Over the years, this was a pattern that repeated itself again and again, and -- little by little -- Analog's product catalogs began to get thicker.

That same sensitivity to the changing market led to Analog's first acquisition, in 1969. At the time, the minicomputer revolution was gathering momentum, and Stata saw an urgent need for Analog to adjust its product line accordingly. "The same customers who were buying our op-amps needed the ability to translate analog signals into digital signals which computers could read," he recalls. Rather than develop the technology from within, he decided to spend about $1 million for a small company, Pastoriza Electronics Inc., that had pioneered in the development of analog-to-digital conversion devices.

That year brought another decision with even more far-reaching implications: the move into manufacturing integrated circuits. Again, Stata looked at changes in technology -- this time, the trend toward smaller and more complex electronic components. His conclusion was that Analog would have to become a player in the emerging market for silicon chips, or it would almost certainly pay a price in future growth. The capital investment would be in the millions of dollars, a huge bite for the company, which had annual sales of about $9 million at the time. But Stata considered the move so important that he was determined to raise the capital one way or another -- and eventually did so through an initial public offering.

To enter the semiconductor business, Analog took an unusual approach that, in some respects, foreshadowed its future experiment with Jupiter Systems and the other participants in the Monterey conference. Instead of developing the new capability from within or acquiring an existing company, Stata worked out a deal with a talented team of young engineers from another Massachusetts company. "They had the technology, and we gave them the capital to form a new company," Stata recalls. The new company was called Nova Devices Inc. It started out with initial equity of about $2 million -- and the use of Analog's aggressive marketing organization. In addition, Stata negotiated with Nova's founders for options to buy the rest of the business as it achieved success in an emerging new field.

"Our goal was to become a market leader," says Stata. "So we figured that the least risky approach was to get some of the best people in the world working on our team." The gamble paid off. Analog completed its purchase of Nova Devices in 1971. In 1973, its semiconductors represented about 25% of sales totaling $22 million.

In the late 1970s, two major areas of concern began to converge, both of which affected Analog's ability to identify and react to changes in technology. On the one hand, the company was finding it harder and harder to finance the costs of research and development without undermining the expansion of the more established product areas. Then again, even when the capital was available, management had trouble recruiting and training all the different experts it needed.

As far as the capital problem was concerned, the situation was aggravated by the sorry state of the stock market at the time, which pretty much ruled out a secondary public offering. "The stock market was essentially dead in 1976 and '77," recalls senior vice-president Larry Sullivan. So, in the summer of 1977, Analog worked out an equity deal with Standard Oil, which invested $5 million, receiving 15% of Analog's stock in exchange.

The shortage of trained experts, however, posed an altogether different sort of problem -- one, moreover, that money aIone could not solve. It manifested itself in Analog's apparent inability to develop certain products -- for example, a new line of electronic converters that could process high-frequency video information for use in CAT scanners and other high-speed instrumentation. Twice, the company had attempted to develop the expertise internally, and twice it had failed. "We were discovering that there was a substantial gap between what we knew and what we needed to know," says Stata.

Eventually, Analog solved that particular problem by acquiring, in late 1978, a small ($3 million a year) company in North Carolina named Computer Labs Inc. But as Stata and his top managers began to formulate plans for the 1980s and beyond, they realized that this was only one in a growing list of new areas in which Analog needed comparable expertise.

Not that the company was in any immediate trouble. On the contrary, it appeared to be at the peak of health, with annual sales fast approaching the $100-million mark. Within a few years, however, various breakthroughs in technology -- the introduction of very-large-scale integrated circuits and the like -- would dramatically alter the marketplace, reducing the cost of processing systems and thereby inhibiting the growth of Analog's basic business in assembled components and linear chips. Already some customers were beginning to demand more than just components. Down the road, as the cost of powerful circuits plummeted, more and more customers would be looking for complete systems to process analog information -- systems that would, in effect, allow them to automate their laboratories and factories.

"The lines between systems and components were beginning to blur," says Stata. "If we saw life as only a component company, we knew we'd be losing out."

During the winter of 1979-80, a task force of Analog's top executives and engineers got together for a few hours every Saturday morning to map out an approach to the challenges ahead. They soon realized that there were all kinds of fields the company should move into, or at least know about -- from advanced chip design, to peripheral devices, to microcomputer engineering. Unfortunately, Analog did not have the resources to pursue so many things at once. "All of our cash and all of our people -- even new people we would hire -- were going to be tied up expanding the businesses we were already in," Stata notes. "And there were some important internal diversifications which we needed to support as well." Yet Analog could not afford to ignore the coming changes in the market. "We knew we needed to cross a lot of new frontiers so we could maintain parity with competitors."

In an effort to find a solution, Stata and Sullivan flew to Chicago in June of 1980 for a meeting with executives from Standard Oil, to whom they explained their predicament. "They told us about the areas they felt they should be moving into," recalls Gordon C. McKeague, president of Amoco Technologies Co., a subsidiary of Standard Oil. "And the vision they had for the electronics industry seemed to make a lot of sense."

Later that summer, Standard Oil agreed to collaborate with Analog in an unusual type of joint venture. Analog would set up a new subsidiary, called Analog Devices Enterprises (ADE), which would make investments in small companies in important areas of technology. The money for the investments, up to $10 million each year, would be furnished by Standard Oil, which would receive preferred stock convertible after five years to common shares in Analog, based on the amount of investment money advanced to the new subsidiary.

"We now had an extra source of money," Stata notes. The next step was to locate the companies to invest in.

ADE got started in the fall of 1980 much as any new venture capital firm would: With money and an eye for talent, it began looking for deals in promising niches of technology. But unlike a venture firm, ADE had goals that were not strictly financial. "Analog had its own diversification shopping list," says Sullivan, who headed up the effort. "We were looking for situations that seemed to smell right."

The first such situation presented itself in the form of a team of Utah engineers, who approached Analog for seed capital to produce digital signal processing circuits. "We were familiar with the field," recalls Bob Boole, ADE's director of venture analysis and Analog's former head of corporate marketing. "What they had, and what we wanted, was a different process for making high-speed circuits." Sullivan and Boole studied the proposal and ran it by an advisory panel made up of Analog's own marketing and engineering people -- whereupon they wasted no time in deciding to invest $700,000. The agreement left open the possibility that the Utah group might eventually sell the company, called Signal Processing Circuits Inc. (SPCI), to Analog. (In fact, they did just that.)

With one investment in hand, Sullivan and Boole continued to search for opportunities in other areas of importance to Analog's future. They were particularly interested in the new field of array processors, peripheral devices used for boosting the number-crunching capabilities of minicomputers. Such devices had many potential applications for Analog's customers. They could, for instance, be used to enhance computer-aided design by making it possible to test a new product's structural soundness almost instantly, in so-called real time.

For more than a year, Boole and Sullivan scoured the East and West Coasts for the right company. "We talked to everyone who would talk to us -- in all, about eight or nine different possibilities, including possible start-ups," says Boole. Along the way, they learned a lot about the emerging industry and its players, but "we had a hard time finding the right mix of product and people."

Then, one day, they noticed a product advertisement placed by a company called CNR Inc. in a technical publication. Through it, Boole contacted Peter Alexander, the architect of an unusually fast array processor. ADE agreed to invest $1.5 million -- to pay CNR for the rights to the design and to capitalize the new business formed by Alexander. In return, Analog received a modest share of equity in the new company, called Numerix Corp.

Another area that seemed to hold promise for Analog was digital image processing, a technology that uses computers to. assist in interpreting medical and earth-resources photographs. Early in 1981, Sullivan and Boole set out to locate a vehicle for participating in this field, and soon stumbled on to a whole nest of image-processing specialists in and around Silicon Valley. On an impulse, Boole decided to phone Richard Ashcroft, president of International Imaging Systems Inc. (I<2>S), in Milpitas, Calif., to see if Ashcroft had any interest in talking with him.

At the time, I<2>S was six years old, with $6-million-a-year in revenues and a passel of cash problems. Ashcroft, a marketing man with a background in engineering, was spending much of his time exploring options with venture capitalists. "We had known Analog [only] as a component vendor," he says. But the more he learned about Analog's entrepreneurial operating style and its ambitions for the future, the more intrigued he became with the possibility of developing a long-term relationship. "They could give us money and advice like a good venture capitalist, and they could also provide technical support for our products and the resources of a large company."

In June of 1981, ADE invested $3 million in I<2>S, receiving about 45% of its equity in return. "Dick Ashcroft and his people helped us understand the potential for the business," says Boole. "We knew about the current applications for image processing. But they made us aware of whole new areas which we could help them develop in the future."

Indeed, the search for investment opportunities was taking them in a variety of unexpected directions, leading them to technological areas of which they had been unaware. They discovered, for instance, that the users of array processors had very specialized software needs, and that meeting those needs would help the market develop more fully. The insight prompted Analog's decision, in late 1982, to invest $500,000 in Quantitative Technology Corp., a new company in Portland, Ore. "When we started out, we had no idea that software was such an important concern," explains Sullivan. "But after a year in the business, we were getting smarter."

Through its exposure to image processing, Analog also had learned of the importance of high-resolution computer graphics for sophisticated industrial applications. So Boole began looking at and contacting companies in the field. Then, in the summer of 1982, he attended a computer graphics trade show in Boston, where he encountered Jupiter Systems. "Bob Boole saw our product demonstration and he was very up-front," recalls Jon Peddie. "He said Analog was interested in graphics and was making selective investments in related technologies."

As it happened, Jupiter had nearly exhausted the start-up capital from its founders, and Peddie was already in touch with a number of financing sources. As he explored the various possibilities, he became more and more interested in ADE. Sullivan and Boole had many years of marketing and financial experience between them, and they understood the dynamics of building a technology company and the difficulty of achieving success in any market. Moreover,in contrast to many of the venture capitalists Peddie had met, they were forthright about Analog's corporate objectives, without being heavy-handed or arrogant. "Most venture capitalists come in and tell you how much money you need," says Peddie. "They have their sights set on taking the company public. [Sullivan and Boole] made it clear that they might be interested in acquiring us some day if everything worked out, but we, as the founders, would always be free to change our minds if it didn't feel right. Nobody would pressure us to do something we weren't comfortable with. They seemed more interested in growing a tree than a weed." The discussions culminated in a $2-million deal in June 1983, which gave ADE about a fifth of Jupiter's equity and an opportunity to buy more in the future if Jupiter decides to give up more equity.

Meanwhile, ADE was also making investments in companies it had no intention of ever acquiring. Since 1982, for example, it has put $3 million into Charles River Data Systems Inc., of Framingham, Mass. The company, founded in 1973, had recently begun making specialized microcomputers for scientists and engineers. Its first such computer was designed around a powerful new microprocessor -- Motorola Inc.'s 68000 chip -- that gave it the processing capabilities of much costlier systems. Charles River's young founder and president, Rick Shapiro, had made plain his intention to take his company public someday. He was therefore not the least bit interested in selling out to Analog, or anyone else for that matter. On the other hand, ADE could not ignore Charles River's technology, which was bound to affect Analog's traditional market. "We knew that this type of computer would have an impact on our business," says Sullivan. "So we decided to get involved with Charles River Data simply to learn about where the industry was going."

ADE took a similar approach to GigaBit Logic, a Newbury Park, Calif., start-up that was developing new, high-speed chips made of a substance called gallium arsenide for applications in computers. Like Shapiro at Charles River Data, GigaBit founder and president Fred Blum wanted his company to remain independent. Analog respected that position. "We talked to them about their objectives, and we came up with an arrangement which had something for both of us," says Sullivan. ADE agreed to invest $1.5 million to help GigaBit gear up for production. In addition to equity, Analog obtained the rights to market some of the new gallium arsenide semiconductors as soon as they became available and, in the future, to produce its own semiconductors in noncompetitive products using GigaBit's proprietary technology.

Whatever Analog's long-term goals were with regard to specific companies, it was committed to making the relationships work. And it still is. To some degree, after all, they share a common future. So it is in Analog's own interest to see that these companies thrive.

Toward that end, Sullivan and Boole have joined the boards of directors of several portfolio companies, and have arranged for other Analog executives to do likewise. At meetings and in telephone conversations, they offer ideas on subjects ranging from how to buy test equipment to how to hire key personnel. "We make comments, and we try to give the entrepreneurs the resources they need," explains Sullivan. "But we don't tell them what to do. It's up to them to run the business. If it doesn't make sense, they won't do it."

When specific problems arise, Sullivan and Boole often refer the entrepreneurs to experts at Analog. "There's a tremendous amount of experience within that company," says Jupiter's Peddie. "Even in areas where they haven't been totally successful, they've learned lessons that they're willing to share."

In addition, Analog has encouraged the companies to collaborate. As young businesses in related technologies, they frequently face similar challenges. Last summer, for example, both I<2>S and Numerix wanted to expand their marketing efforts in the oil industry, but neither could justify the expense of hiring a full-time, Houston-based salesperson. "I suggested that they talk to each other," says Sullivan, a director of both companies. They did -- and decided to get together and hire a salesperson between them.

It was the desire to explore futher synergy that led Analog to call the Monterey meeting last fall. And, in fact, as the entrepreneurs discussed their businesses over a two-day period, they began to discover a variety of overlapping interests. There was talk of other joint sales and marketing efforts and of combining on volume purchases of components, such as memory chips, to qualify for lower rates. Some of the companies even explored the possibility of buying from and selling to one another. And all of them traded ideas on ways to motivate and reward their employees as their businesses grow in the years ahead. "Monterey provided a forum for everyone to get to know one another," says Sullivan, who helped host the event. "And there was an awful lot of mixing."

The October meeting, however, offered no clear answers to the fundamental questions at the heart of Analog's experiment. How much independence, for example, will the portfolio companies have if and when they are acquired? And what incentives can Analog provide to encourage the people who created the companies to stay on as part of a much larger business organization?

Analog is still searching for answers to these questions. During recent visits to I<2>S and other affiliated companies, Stata has been talking with groups of employees about the importance of mutual benefits in any successful relationship. Analog would be foolish, he contends, to exercise an option to buy any portfolio company without the full backing of its people. "We're not interested in forcing a shotgun marriage on anyone. The onus is really on us to create the type of environment that people want to be part of."

As for Analog itself, Stata has no illusions about the challenges ahead. "We don't know what will happen as we grow from $200 million to $500 million and larger -- whether we'll be able to fight off rigor mortis and bureaucratic diseases. But through all of these affiliations, we're trying as hard as we can to be a company that maintains the vitality and flexibility of a much smaller business."

It will, of course, be years before the results are in. In the meantime, the mere fact that Analog is trying bears witness to the strength of its entrepreneuriaI spirit. "This is really unploughed ground," says Stata. "And if we succeed, we will have untied a Gordian knot."