In December 1975, two months after he left General Automation, Ray became a consultant for System Industries, a manufacturer of disk storage systems based in Sunnyvale, California. The company had been founded five years earlier by Edwin V. W. Zschau, whom Ray described as “perhaps the best educated man in the Bay area, a doctorate in both business administration as well as in mathematics and physics”. Zschau was Assistant Professor of Management Science at the Stanford University Graduate School of Business from 1964 to 1969. In 1970, with financial backing from three venture capital firms (Brentwood Associates, DSV Associates, and California Northwest Fund), Zschau set up Systems Industries. In March 1971, the company shipped its first disk storage system.
The central processing units of early minicomputers had very limited memories—a 1969 minicomputer may have had 4,000 bytes of random access memory (RAM) compared to billions of bytes in a 2010 PC. This meant that the supporting storage devices, such as tape drives and the recently developed disk drives, had to carry a lot more of the load in typical data processing. Whole programs were rarely put into RAM at the same time. They would be broken into parts that would “overlay” each other (replace each other in RAM) as they were needed. The parts that weren’t being used would be stored on a disk. Software applications such as accounting and word processing programs were routinely handled this way as were the databases containing information. Extensive use of overlaying made disk performance just as important as CPU performance in determining overall system performance.
A disk storage system consisted of a disk controller (printed circuit board); software enhancements (programming commands that worked with the minicomputer operating system); cabling and connecting hardware; and the disk drive. The disk drive might be a removable cartridge or pack or, after 1979, a “Winchester” drive, where the disk is permanently sealed within the drive. (The Winchester name was an allusion to the rifle bullet; the first such drives were called 30-30s while they were being developed at IBM. According to the Smart Computing Encyclopedia online, that was because they featured two spindles with a storage capacity of 30 million characters each.)
Every minicomputer needed at least one disk storage system. In the 1970s a significant number of companies were hooking up additional storage systems to the same computer or to multiple computers, to speed access to data and to provide a system back-up should the primary storage system fail. This concept of “system fault tolerance”—compensating for potential faults in the storage system by creating redundant systems, so that the computer can continue operating even in the event of a component failure—also contributed to the demand for storage systems.
The proliferation of minicomputers in the 1960s had created a growing need for high-performance disk storage systems. Zschau’s idea was to bring technological advances in storage systems to the market faster than the competition. When Systems Industries shipped its first product—a system incorporating a new 2.5MB disk cartridge drive—to users of DEC and Data General minicomputers in March 1971, it was in fact several months prior to the release of similar products by DEC and Data General. In 1974 the company shipped storage systems featuring 80MB storage module drives—four years before DEC shipped its similar product.
System Industries started out as a “pilot fish” company swimming around giant minicomputer manufacturers and feeding on their leftovers. DEC and Data General were two of the largest minicomputer manufacturers at the time; they accounted for perhaps half of the US market. System Industries designed its disk storage systems to work with those computers; it also supported military computers designed by Sperry Corp’s Univac Division.
Virtually all the company’s sales until 1977 were to OEM customers: Manufacturers making computer systems for oil and gas exploration, typesetting operations for newspapers and publishers, office data processing and word processing, engineering operations, military applications, and other process automation markets. System Industries was itself an OEM because it bought its disk drives from other vendors. The only components the company designed and produced itself were disk controllers, cabling and connections, and software.
“Purely by luck,” said Ray, “I became acquainted with the company and was invited to come in and see what I could do to help out, because they were shipping about $6 million and losing about $2 million. That’s a difficult thing to deal with particularly when there are emotions involved.
“But together we worked the company out of that situation and back into a growth mode. When I left them in 1980 the company was doing about $80 million in annualized sales and profitable enough that the investors decided they could take it public with pride.”
For just over a year, December 1975 to January 1977, Ray worked as a consultant to the company. In 1976 he was offered a place on the board of directors. In January 1977, he was named Chief Operating Officer. He was subsequently elected President, then President and Chief Executive Officer.
Following a pattern established at GE and General Automation, Ray simultaneously attacked sales, manufacturing, shipping, and expenses. He set higher quotas and spent hours on the telephone making deals. He paced the plant and rode the manufacturing people. He looked over the shoulders of the shipping guys to make sure the product got out the door. He cut budgets, personally reviewed expense reports, and worked to stanch the flow of money. He set a productivity goal in terms of dollars of sales per employee. In 1980, the company achieved $131,000 per employee.
Ray also saw that System Industries’ focus on OEM sales was a barrier to growth. The OEM market was too small and too dependent on the ultimate end user markets served by a few key customers. If System Industries was to achieve sales in the $100-million range, it would have to sell its products directly to end users. Ray made plans to change the marketing direction of the company, and in 1976 development began on a new product—the 9400 series—that would appeal directly to the minicomputer end user. The plan also called for an expansion into international markets and creation of a service and support business.
At the end of Ray’s first year (1976), revenues had doubled from about $6 million to $12.1 million. Instead of a net loss of $2 million, the company showed a net income of $703,000. The 162 employees who worked at System Industries felt proud, and relieved, that the company was finally in the black.
In Ray’s second year, System Industries still sold almost exclusively to OEM customers, and it was selling systems that had been developed in 1975 and earlier. Development of the 9400 series—designed to operate with DEC PDP-11 software—was in full swing, with a projected ship date of first quarter 1978. At the end of Ray’s second year (1977), revenues were up 35%, from $12.1 million to $16.4 million. Net income was up 70% to $1.2 million.
Ray’s third year was a rough one. As the company built a direct sales force in the US and Europe and expanded its customer service capabilities, technical difficulties caused a delay in the development of the new product line. When the 9400 series finally shipped late in the year, it was nine months overdue. Consequently 1978 sales were up just 14% over the previous year to $18.7 million, and the company posted a net loss of $1.4 million.
In Ray’s final year and a half with the company (1979–80), the engine he had so carefully tuned finally started humming. The transition to an end-user marketing focus was completed. Sales of the 9400 series accounted for nearly half of all system sales and the customer service revenues almost doubled.
The decision was made to phase out an unprofitable subsidiary business, Silonics, that had manufactured an ink-jet printer. System Industries licensed the technology developed by Silonics but no longer manufactured printers. Ray and others decided the company should focus on its primary storage systems business.
Revenues in 1979 were $25 million, up 33.5% over 1978. In 1980, sales increased 51% to $37.9 million. When Ray left the company at the end of July 1980, Systems Industries had an installed base of 11,000 systems and was on track to achieve sales of $63 million in 1981. The company went public in September 1980.
In 1980, Ray received $65,255 in salary from Systems Industries, plus an additional $3,100 in personal benefits. His annualized salary was probably in the $110,000 range (Zschau made $146,899, while two Vice Presidents made $212,782 and $152,076, respectively). Ray seems to have preferred to take most of his compensation in the form of company stock, of which he had conspicuously more than either of the Vice Presidents who drew higher salaries than he. When Ray left System Industries, he had options to purchase 38,250 shares—2.43% of the total shares. Only the founders and the venture capital firms and principals had interests in more shares. Ray’s options were worth an estimated $562,656 as of December 1980.