If you're a struggling young high-tech entrepreneur with a new product design and much more time than money, you might decide not to buy preprinted circuit boards and instead produce your own. "You start with a fiberglass board completely covered with copper," says Casey Cowell, who spent a good deal of time at this in the late 1970s, "and the object is to remove the copper you don't want"--sort of like a sculptor removing from a block of marble everything that is not the statue.
Cowell and his partners drew up a wiring diagram for their computer-communication device, then photographically imprinted its image on a silk screen stretched over a wooden frame, so that the screen would let ink through only where they wanted wiring to run. Laying the screen on a copper-covered board, they would then squeegee a very heavy ink over it--and when all went well, they wound up with a board inked in all the places where the wiring would go, and in only those places. Finally, they dunked the board in an acidic bath that dissolved all the copper not covered by ink. Once they washed off the protective ink, they had a properly wired circuit board ready for its complement of switches, integrated circuits, transistors, resistors, and capacitors.
"It takes only two or three seconds to silk-screen a board once you get everything set up," says Cowell. "You can move right along. You can do several hundred units at a time. We did it in a small apartment." He pauses for a moment, reflecting. "A very small apartment."
Cowell is now chairman and president of U.S. Robotics, a $37 million computer communications company with 200 employees, a 100,000-square-foot Skokie headquarters, and sales subsidiaries in the United Kingdom, Scandinavia, the Netherlands, and Japan, all of which grew out of that apartment near 47th and Lake Park. "If Casey is anything, he's a survivor," says accountant Richard A. Reck, who as partner in charge of KPMG Peat Marwick's high-technology practice in Illinois knows well the industry's perils. "They lived through a couple of shakeouts that could have killed them either time--but they've survived and are moving forward quite strongly." Adds Nina Klarich, president and CEO of the Chicago Technology Park and president of the Chicago High Technology Association, "I just enjoy the whimsical eye with which Casey views his own success."
It's been more than a decade since Cowell and his partners (Paul Collard, who is no longer with the company, and Steve Muka, who died in 1985) laboriously created their own circuit boards--as well as practically everything else that went into their product--but he remembers the experience well. It dissolved most of the preconceptions he began with, leaving behind two pieces of advice that he urges on today's generation of novice entrepreneurs:
(1) Chicago is a great place to start a technology-based business.
(2) Don't start it the way I did!
Casey Cowell graduated from the University of Chicago in 1975 with a bachelor's degree in economics. Like most of his classmates, he went right to graduate school--in his case to the University of Rochester, with the aim of becoming an economics professor. A friend used to needle him, saying that once Cowell got his PhD he'd be the only person in the unemployment line who knew exactly why he was there. And Cowell soon came to agree: he didn't want to be in school anymore. "For me, graduate school was really far removed from participating in the mainstream of commerce."
Not that the mainstream of commerce was anywhere in sight at Christmastime 1975: "I was driving a beat-up Volkswagen with no heater and holes in the floorboard back to Chicago, wondering, is this my real calling? What are my options? It's pretty humble, and your planning horizon is about one day."
Cowell had never seen the professoriat as his only possible future. He'd grown up in Detroit, where his father had been an independent heating and refrigeration contractor. An older friend down the block had started a couple of businesses Cowell worked for. All this may have immunized him against the anticorporate campus climate of the early 70s; at any rate he knew from his own experience that real people did run businesses. "For me, at that age and time, starting a business was a real legitimate, valuable thing to do. That option was open to me."
At the end of the school year he moved back to Chicago--"it was a foregone conclusion in my mind that I should start where I knew people"--and joined up with four technically minded fellow U. of C. graduates (two soon departed) who agreed with him that computers and electronics sounded like an exciting area to get into.
They pooled their money--$200--and all the time they could spare from miscellaneous full- and part-time jobs. They never really planned to build robots, but one of the group was a fan of Isaac Asimov's 1950 science fiction classic I, Robot, which described a company called U.S. Robot and Mechanical Men, Inc. So they dubbed their not-yet-fledgling business "U.S. Robotics."
You have to remember that in 1976 a pocket calculator was cause for delight. Legendary Silicon Valley start-ups like Apple Computer did not yet exist. Laughs Cowell, "Back then, if you said 'robotics' people couldn't spell it, and if you spelled it they couldn't say it. But once you explained it to people, it conveyed technology in general, which was what we wanted."
You can't sell "technology in general," though. Cowell and his partners first set out to build a computer keyboard with a built-in acoustic coupler. "We got the acoustic coupler part done first, and decided we'd sell that."
An acoustic coupler was a device enabling computers to exchange information over a distance. You'd dial the phone number of another computer and place the handset in a pair of oversize rubber cups wired to your computer. Then the acoustic coupler would translate the computer's electronic pulses into audio tones, the telephone would translate the audio tones back into electronic pulses for transmission, and then at the other end the reverse process would take place. This was all necessary because it wasn't until 1976 that the FCC wrote regulations allowing products not made by AT&T to be plugged directly into telephone lines (today's modems, of course, do just that); this quadruple translation process made communications relatively slow and prone to error.
USR's first acoustic coupler--"a real Model T," grins Cowell--could send the information on a double-spaced typewritten page from one computer to another in about a minute at 30 characters per second. Today a top-of-the-line USR modem does it in two-thirds of a second at 2,600 characters a second, saving not only time but telephone bills. Come to think of it, says Cowell, the old model was more like "a Model T that you put your suitcase in and then got behind it and pushed."
Cowell was 23 years old. "We were long on enthusiasm and short on experience," he recalls. "It was probably the time in my life when I had the fewest restrictions. We were willing to put in endless hours and work very hard, and we didn't know any better. No one had a list of reasons why they couldn't work all hours. The motto was, sleep faster."
They decided to manufacture the couplers themselves because it seemed cheaper. "If you went into rebuilding car engines," says Cowell now, "you wouldn't design and manufacture every part of the assembly yourself. But that's what we did. If this had been a more mature industry at the time, it would have been very hard to compete on a do-it-yourself basis." But it wasn't, and they did. "We had access to a machine shop. The units were to have a plastic case, which was made through vacuum forming with a mahogany mold. So we bought blocks of mahogany and learned to machine them into the proper shape for the mold.
"And that's one of the neat things about Chicago. Every conceivable manufacturing process gets done here. We found a guy with a plastic molding company who was willing to talk with us and explain how to make our next block of mahogany work better as a mold. You could drive in your car to people like that--it's not like there were just one or two plastics molding shops, they're all over the place, so you can find someone willing to help you.
"We thought it should have a plastic nameplate--so instead of buying one [ready-made], we silk-screened it ourselves. We made our own circuit boards.
"We bought our base plates from Ryerson Steel on the near southwest side. I'd drive my red VW into this enormous warehouse, and get in line with a bunch of semis, underneath these big cranes big enough to pick up my car--and load up with two boxes! Some people thought it was a little odd--but after a while they would be cheering for us because we'd ordered three boxes instead of two. . . . It was a mission."
But not a mission he would encourage anyone else to undertake. "I would absolutely not recommend that people do it the way we did. If I were doing it over, I would work for a while at a couple of companies in the field, maybe a big one and a smaller one, in order to get a couple of things I wish I had had that I didn't: sales and marketing experience, and role models." Especially role models: "As a business grows and gets into new terrain, it's often useful to envision how somebody else would handle a situation like yours. The disadvantage of starting without experience is you don't have that.
"A more typical way to start"--now--"would be to see an opportunity to manufacture data-communication products for people who want to call other computers," says Cowell. "And you'd talk to lots of those people and find out what they want and where they buy it; figure out what it would take to design and manufacture the products"--i.e., devise a business plan--"raise the money and spend months or years designing and developing the product, and getting marketing and distribution set up. At the end, you'd have a very focused business and if you worked hard and had a little luck, you might succeed.
"Instead, we said, "This seems like a neat industry to be in, we have $200 cash and side jobs to keep us eating, so let's go to work on this neat product idea ourselves.' In a more mature industry, it would have been very hard to compete in that way. There should be more productive uses of your resources and expertise than soldering components or silk-screening circuit boards."
Richard Reck of Peat Marwick agrees. "Back then you could afford to make mistakes. They were probably spending twice as much on their product as necessary. But margins were larger then."
Now that margins are narrow, entry barriers high, and competition tough, business skills are at a premium, even in high-tech businesses. Says Tom Parkinson, executive director of the five-year-old nonprofit Evanston Business Investment Corporation, "We have learned that the most important thing [about a start-up business] is not the product but the management and marketing skills. It's easier to make a mediocre product successful with good marketing and management than it is to make a good product successful with mediocre marketing and management. The more successful high-tech entrepreneurs are people who have their business plan and marketing strategy even before their product is completed. On the other hand, inventors--people who have a neat product already and only now want to put together a business plan--usually face the most obstacles and usually are the least likely to succeed." No wonder Cowell says do as I say, not as I did!
U.S. Robotics survived, partly on spunk and sweat, partly on lucky timing. The 1976 FCC decision allowing independent manufacturers (who met technical standards) to make devices that plugged into the telephone network allowed acoustic couplers to evolve into all-electronic modems. Even more fortunately, the computer industry as a whole was economically "immature," a kind of undifferentiated blob: there were few competitors (now Cowell estimates there are about 100) and no big, established modem manufacturers; there were no entrenched channels of distribution, no exclusive franchises, no barriers between the enthusiastic designers/inventors/manufacturers/distributors in Hyde Park and their equally enthusiastic customers. USR sold its first modems through page-nine ads in Byte magazine and got along fine. Compare life then to Cowell's experience last year, when USR arranged a large potential deal with a Fortune 500 company, which, however, makes all its computer purchases through a particular dealer chain that had never sold USR modems. "We put a done deal on the table which represented a significant amount of revenue and profit for [the dealer] selling to a significant customer of theirs. They were not interested," says Cowell, "because it would cause them to add another product to their product line."
The company shipped out its first modems in January 1978. "We had I think $50,000 in sales that year. We actually paid ourselves some money. The idea that the business might pay us something was a real novel idea at the time--something we had thought was still off in the future."
The rest of the industry was growing, too. Before the middle 1970s, people who bought computers usually bought terminals from the same outfit, but then independent manufacturers also began producing them. "People all around the country liked our modems," says Cowell, "and they would ask us which terminal we recommended. After this happened often enough, one day the light went on and we decided to distribute terminals manufactured by others, as well as our own modems.
"Distribution became the bulk of our business. We sold products for Televideo, DEC, GE, Perkin-Elmer, Applied Digital Data Systems, Teletype, and others."
This move kept money coming in and enabled USR to survive and grow. But it could easily have been the first step into economic quicksand. "Many small businesses do well at the start," reflects Cowell now, "but they are extremely opportunistic," jumping into every vaguely related line of work that comes along. "You can end up doing a mediocre job of several unrelated things. Your company will generate more synergies if it's more focused." And going into distribution was a very opportunistic move.
Today, almost the first thing you see when you come in the company's front door is its mission statement, adopted in 1988: "U.S. Robotics, through total employee participation, will be the best data communications company, providing quality products, services, systems, and support to meet the needs and expand the capabilities of business and professional customers worldwide." If some such statement had been in the forefront of the partners' minds in 1978, "we could have gone to other companies selling terminals and said to them, 'We want to build modems for you that you can sell with your product,'" thus keeping their focus on what they did best.
Distributing others' products may have fuzzed up the young company's focus, but it did prove to be educational. "When the products didn't work, the users called us up directly. We learned all the reasons people had problems, what they liked and what they didn't, how various terminals worked with our modems."
From this feedback they got the idea of sticking a label with all the basic operating commands on the bottom of each modem--"it's easier than having to call someone up or scurry around looking for the manual." For some users, it turned out that desktop clutter was a real problem, so USR brought out Password ("the urban jeep of modems," says Cowell) in 1983. It came complete with its own double-headed connector cable and a patch of Velcro on the bottom so that it could be stuck on some vertical surface out of the way.
By 1981, U.S. Robotics had moved from Hyde Park to Lincoln Park and then back south to Lake Street and then to Washington Street a few blocks west of Halsted. About three-quarters of its business was selling other people's terminals and equipment, about one-quarter selling its own modems. "We had five years' experience and a going business," says Cowell, "enough tangible and intangible resources, and enough perspective to say, 'Is this what we really want to do?'
"We were really in two different businesses--distribution and manufacturing--and we had to choose our focus. We could have continued to be a distributor, but we decided we wanted to be a designer and manufacturer of data communications products, focusing on the small-computer marketplace." (Five years before, of course, there was no small-computer marketplace, so that decision couldn't have been made then.)
Easier to say than to implement. "It was definitely a character-building year. We had been selling direct to end users. Now we wanted to sell our data-communications products through distribution companies we had just been competing with!
"We had to completely reposition ourselves, go and tell them, 'You know us, we used to be your worst enemy. Now we want to be your best friend--but not for six to eight months, while we get reorganized.'"
The repositioning worked, a new line of products sold well, and U.S. Robotics for the first time went to professional investors for more money to help them grow. In 1983 the partners also decided to go after the fast-growing "OEM" market--designing and making modems for computer companies ("Original Equipment Manufacturers") to then sell under their own brand names. "We were successful in landing a number of fairly large contracts--with Apple, Kaypro, Commodore, Zenith, Grid, and Televideo," says Cowell.
But sudden success can be as dangerous as sudden failure. The company's building at 1123 W. Washington had never been altogether satisfactory. Reck recalls, "Their back door wasn't big enough for them to palletize the loads for shipment. They had to load their boxes by hand onto a pallet on the truck." Now that orders were rolling in, the building itself was too small.
One of Cowell's favorite things about Chicago is that it's easy to get anywhere else from here. For essentially the same reason he says he would happily have stayed in the near-Loop west side. But he just couldn't find anything big enough at a reasonable price. "When the orders start coming in, you don't want to say, 'Yeah, we'd like to have it for you, but could you wait a couple of years while we build a new building?'" By the end of 1984 they were on McCormick Boulevard in Skokie.
In 1984, sales to the big boys made up well over half of U.S. Robotics' business. But the contracts started drying up when the personal-computer business slumped in 1985. USR had been selling to both consumers and OEMs; now half their business was vanishing. Like it or not, it was time for another strategic course correction.
Cowell discovered, once again, that he was in two different businesses at once. "In the OEM business, you have a few large customers and you do what they want. Your sales organization and everything else is geared to that. In brand-name sales [to consumers], you create an organization good at finding where the market is headed, and develop products accordingly. That means you need a different sales expertise. Your research-and-development group has to be different, and interact more with marketing"--since you're trying to follow the market yourself, not follow someone else who's following the market.
"So they're really two different animals. It's hard to run an organization that's trying to do two different things. You're more likely to succeed if you're doing a few things well rather than several things in a mediocre way."
Only a few years before, there had been no room in the computer business for such fine distinctions. "In the early days," says Cowell, "it was common for any wholesaler to have a retail store or stores and run a mail-order company. But over time, as the industry reaches a certain size, particular segments are big enough that specialization pays. A wholesaler [who is just a wholesaler] can be better because that's all he does--there are no mixed signals, no conflict." Since 1985, USR has focused on selling its own brand-name products; OEM sales are down to about 5 percent of the business.
With all this talk about focus, I was surprised to learn that U.S. Robotics just last month bought a software company in Baton Rouge, Louisiana. Modems are hardware, after all; they've been the company's bread and butter for years. But Cowell says, "It's a very natural fit for us. If you have an IBM PC or a Mac, and you want to use it to send information over the phone lines, you need both a modem and a communications software package of some kind to manage the computer to use the modem.
"With this acquisition, we can move from selling parts of that solution to selling all of it and doing it in better ways." His point is that there is more than one way to improve the product. One way is to keep on building faster and faster modems. But equally important is to find new and unexpected uses of the existing technology--much as early moviemakers learned how to do more than just fix a camera in one spot and film a stage play. In the case of modems, such innovations are likely to require new programs. So, "having a software company enables you to do that second kind of improvement much better."
Cowell remembers only one other bootstrapping entrepreneur from the late 1970s--a fellow UC alum who developed a program called Typing Tutor that ran on Apples--and he doesn't know what became of him. Cowell and his partners did not hang out with other people who were trying to build a new technology-based business. For one thing, they were too busy; for another, they didn't know any. "We were usually met with total skepticism, even in the neighborhood. It's a good thing there were a few of us. One person by himself could have become discouraged."
And if federal, state, or local government was trying to help high-tech ventures, Cowell and company knew nothing of it. Until their 1984 move to Skokie--when the city of Chicago tried and failed to help them find enough space to stay in town--Cowell can't think of any government action, at any level, that either helped or harmed his business. "You tend not to think about it."
Like starting a business with $200 and enthusiasm, this kind of isolation may have been tolerable in 1976, but in the more competitive, more structured world of 1990 it's a serious handicap. Yet until quite recently it was the rule. Young technology-based enterprises like U.S. Robotics have had few sources of help or information, few people to talk to, few ways to find (for instance) an accountant or lawyer who would understand their particular needs. "I don't remember how exactly we found our original lawyer and accounting firm," says Cowell, "but it was basically through a network that wasn't very focused on one thing--just like calling up people you know. There were no three businesspeople I could call up and just ask, 'Who is your lawyer?'"
In a word, Chicago had no high-technology network until the middle 1980s. It's hard to understand why, in a metropolitan area littered with good universities, federal and corporate research laboratories, giant electronics manufacturers, and--according to Ed Grant, executive director of the local Information Industry Council--the largest installed computer capacity in the United States. Evidently, even in the minds of those who should know better, the stereotype of Chicago as a heavy-industry, mammoth-corporation town dies hard. (Nor does the city give successful entrepreneurs the recognition some think they deserve. "If this were California," says Grant of Casey Cowell, "they'd be carrying him around in a chariot.")
Nevertheless, a healthy high-tech network is spreading. The Chicago High Technology Association organized in 1984. The comparatively venerable American Electronics Association (born in 1943) finally established a midwest council (headquartered in Palatine) in 1988. "It's the world's largest electronics trade association," says Cowell, who has been active in it--"a great group to plug into and meet people." The Information Industry Council--itself an offshoot of the Civic Committee of the Commercial Club of Chicago--is about to publish a detailed directory of area software companies, something Grant says no other metropolitan area has. Incubators like the Chicago Technology Park--where fledgling companies can share support services and business woes over coffee--are no longer novelties. Inventors and investors can find each other through groups like the Inventors' Council and the MIT Enterprise Forum.
To point up successful role models, Richard Reck helped start Peat Marwick's Illinois High-Tech Entrepreneur Awards program in 1983. Cowell was one of the 1984 winners; the 1989 awards went to Cyborg Systems (Chicago, specialized personnel-management and accounting software), Appliance Control Technology (Addison, electronic appliance controls), Resource Information Management Systems (Naperville, health care software), and Frasca International (Champaign, flight simulators). Reck believes Chicago's high-tech rocket is poised for takeoff. His own accounting specialty started with 26 Chicago-area clients in 1982 and he now has about 200. "Every week or so I run into a brand-new entrepreneurial situation."
Groups like CHTA and AEA "are an ongoing catalyst," says Cowell. "They create a word-of-mouth network among technology industries, so that a certain amount of collective wisdom develops. Everything else being equal, people are more likely to succeed when that happens." Or as Reck puts it, "The guy who's about to make the mistakes Casey made can get hold of him now."
Art accompanying story in printed newspaper (not available in this archive): photos/John Sundlof.