Father, Son & Co.

by Thomas J. Watson

  The Williamsburg organization came none too soon, because IBM’s growth was accelerating. The business grew faster in the two years after Dad’s death than ever in its history, except the wartime expansion of 1943. Once we won the numbers race against Remington Rand, the psychology of the market swung completely in our favor. Even though computers were the most complicated and expensive business tools anyone had ever seen, customers were deciding they couldn’t be without them. They picked IBM because of our reputation for both machines and service. We had the salesmen and technicians who knew how to provide systems that worked, scores of experts who could unravel tough programming problems, and a huge library of computer programs we offered to customers at no charge. Getting a computer was a major investment that required approval by the board of directors in most companies, and for the executive in charge of picking the right machine, IBM became known as the safe choice. As Fortune magazine put it at the time, “Boards of directors may know little about machinery, but they know about IBM.”

  Although we were having success after success, I worried about whether we could hang onto the advantage we’d gained. RCA was coming into the industry, and I thought they were going to be terribly tough. On the other hand, Al Williams always believed the most serious competition would come from General Electric, which around that time landed a huge contract to computerize the retail banking operations of Bank of America. GE is a smart company, well organized, and when they take hold of something like jet engines or even dishwashers, they really do it thoroughly. In the mid-1950s RCA was half again as big as we were and GE was five times our size. If either company had decided to hire away some of our best people and commit large amounts of money to this thing, they would have wiped us off the map.

  Shortly after Dad died, while we were still planning for Williamsburg, I was summoned over to RCA to see General David Sarnoff. RCA had been experimenting with computers since the 1940s and had just delivered a giant machine called BIZMAC to the U.S. tank arsenal in Detroit, where it was used to keep track of inventories. Some RCA executives believed computers would be as important to their company’s future as color televisions. It was hard for me not to feel awe standing in front of the father of American radio and television. The general was a short man, but he sat either on a high chair or on a platform and he looked very big to me. He had a long cigar that he pulled at sporadically as he spoke.

  Although I had met him socially, this was a different General Sarnoff. He said he wanted licenses under our computer patents and he thought we were reluctant to make them available, even though we were required to do so under our new consent decree. He told me that the expertise that produced RCA’s TV sets could also be used for computers and that he intended to capitalize on that. I didn’t let my feeling of awe stop me from pointing out that computers were a highly specialized market, more dependent on sales and systems expertise than hardware. I said that while we would license him any patents he wanted, I thought he would have a difficult time in the computer business.

  Meanwhile one of the people helping us get ready for Williamsburg was a management consultant named John L. Burns. I knew Burns casually—a big, heavyset Harvard Ph.D. who had started his career during the Depression as a laborer in a steel mill and had risen to senior partner in the consulting firm of Booz Allen & Hamilton. When I’d asked him to work with IBM, Burns had explained he was a top consultant to RCA and would have to get their permission. Then he accepted and worked quite intimately with us until Williamsburg. Not more than three months after the conference, Burns telephoned and said that Sarnoff was offering to make him president of RCA. He asked if I had any objection. I said, “I most certainly do, John!” because we had entrusted him with detailed knowledge of our organization and methods and plans. Nevertheless he took the job.

  The threat posed by Burns would have been much more grave had I not paid attention to some advice I got from Al Williams. Al had telephoned one day to tell me that Booz Allen’s team was asking for an explanation of IBM’s pricing practices. “Sure,” I said. “It’s like your doctor, you have to tell them everything.” But Williams insisted it was unwise to give them the data. I was irritated because I thought highly of Burns, but I always made a point of listening to Al. So I said, “All right, if you feel that strongly don’t tell them.” The pricing practices had been built up since the early days of IBM. Behind the price of each rental product we had our own cost of marketing, cost of servicing, and rate of planned obsolescence—all carefully kept secrets. In Burns’s hands this knowledge would have enabled RCA to aim where IBM’s product line was weakest and avoid attacking where we were strong and could cut prices to fend off competitors.

  We tried to make sure that for well-financed rivals like GE and RCA, the computer business seemed too risky a bet. We did this by making ourselves tremendously sensitive to toeholds. The feeling among our sales force was that if a General Electric could get five percent, it could get 100 percent. If a salesman lost a customer without having first alerted management that the account was in jeopardy, he was subject to discipline. IBM salesmen had two types of call reports to fill out. The first were routine; the second were named special account reports and came in two colors. Pink ones were for situations where IBM was bidding against somebody else to win a new customer; yellow slips represented sick accounts, where a customer was dissatisfied. These reports were compiled and exhaustively analyzed by geographical location, by customer type, and by product type. Coupled with other research, the special account reports enabled us to assess accurately how the competition was doing.

  Any inroad made by a competitor like RCA was worrisome. In an industry exploding like ours, it was one of my unwavering convictions that gaining and holding market position was crucial. Any deviation from this goal in an attempt to maximize short-term profits would, over the long term, reduce the total amount of our profit. By the same token, to try to pick and choose a limited number of areas in which to be strong was a foolish and dangerous course. If we were to do that, we would limit the scale of our business, making it easier for competitors across the board.

  I believed that the only way for IBM to win was to move, move, move all the time. As the computer industry grew, we had to grow with it, no matter how fast that growth might be. I never varied from the managerial rule that the worst possible thing we could do would be to lie dead in the water with any problem. Solve it, solve it quickly, solve it right or wrong. If you solved it wrong, it would come back and slap you in the face and then you could solve it right. Lying dead in the water and doing nothing is a comfortable alternative because it is without immediate risk, but it is an absolutely fatal way to manage a business. So I never hesitated to intervene if I saw the company getting bogged down. A few months after Williamsburg we were having a dreadful time moving over to transistors. The transistor was obviously the wave of the future in electronics: it was faster than the vacuum tube, generated less heat, and had great potential for miniaturization. Nobody was selling transistorized computers yet, but a great many companies were racing to perfect them, including RCA, Honeywell, Control Data, NCR, and Philco. We too were experimenting with transistorized computers and calculators in our Poughkeepsie lab. Early transistors were unreliable—they were sensitive to heat, moisture, and vibration—but by 1956 Ralph Palmer and the people he directed at Poughkeepsie had done enough testing to be sure those limitations could be overcome.

  The big remaining hurdle was cost. Transistors were selling for about $2.50 apiece and it seemed impossible to design a computer using them that would ever make any money. But Birkenstock, as usual the one to goad me in technology, pointed out that by moving aggressively we could make this difficulty disappear. The leading supplier of transistors in those days was Texas Instruments in Dallas. Not many people had ever heard of them, but they’d stolen a march on big vacuum tube makers like GE and Sylvania by learning to mass-manufacture transistors first. Pat Haggerty, their factory chief and later president, understood th
e unique economics of the transistor game better than anybody else. When Texas Instruments first went into the business, transistors were costing up to sixteen dollars each and one of their few practical applications was in hearing aids. Haggerty figured that by getting the price down to about $2.50, the company could open up a mass market in portable radios. They gambled two million dollars on a circuit design and production process to meet this goal, and in 1956 the transistor radio was born. It became a hit with consumers and put Texas Instruments on top.

  We thought similar magic might work for IBM. By designing computers on the assumption that the cost of transistors could be pushed down still further, to perhaps $1.50, we could set prices that would attract customers and still let us make a profit. Birkenstock flew to Dallas to ask Texas Instruments to help. They agreed to build a factory with high-volume production lines that would cut the cost of transistors so dramatically that they’d be cheaper than the high-grade vacuum tubes we were now depending on. In turn we promised to use the lion’s share of the millions of transistors the new plant would produce.

  That’s where we almost bogged down. This bold plan gave us an incentive to transistorize all our products, computers and punch card machines alike, because the more transistors we used, the cheaper they’d be. While the idea of transistorization delighted the computer engineers at Poughkeepsie, it raised a storm of protest from the punch-card designers up in Endicott. They had barely learned to handle electrons in a tube, and this new invention shocked them. I would go up to the lab and say, “Why not transistors?” hoping they’d take the hint. But for months every new design they sent down to New York was full of tubes. Finally I issued a memorandum that said, “After October 1, we will design no more machines using vacuum tubes. Signed, Tom Watson Jr.” The Endicott people were awfully mad and said, “What does he know about it?” But I got a hundred of those little transistor radios Texas Instruments was making and carried a few along whenever I went to Endicott. Each time I heard an engineer say that transistors were undependable, I would pull a radio out of my bag and challenge him to wear it out.

  On Wall Street, IBM was a darling stock. Brokers in those days used to tout promising new companies as “the next IBM.” The value of the stock had multiplied fivefold just since I’d become president; and if your family had been fortunate enough to invest $2,750 in a hundred shares during the year that my father took over the business, by 1957 that stake would have grown to $2.5 million. A big event on Wall Street that year was our first stock sale. Father had always rejected the idea of selling new shares, but it finally came down to a question of common sense. We’d borrowed as much as you could borrow—we owed the Prudential well over $300 million, which made us the biggest debtor in American business. Yet at the rate we were building factories and rental equipment, we were going to need more capital—a lot of it. Al Williams calculated that we could easily make use of $200 million. So I called our man at Morgan Stanley, Buck Ewing. Up to that moment he was probably the most underemployed investment banker of any major corporation. Buck and I had flown together in World War II, and after the war Morgan Stanley had assigned him to the IBM account. The first time Buck had called to ask if IBM needed capital, I told him, “I don’t think my father is ready for this,” but I introduced him to Dad all the same. They chatted very briefly. After Buck left, Dad said, “That man is not using his head. The last thing we want to do is sell stock.” The head of Morgan Stanley was Perry Hall, whom Dad knew from the Short Hills Episcopal church. Hall was sharp; he had listened to Dad’s views on selling new stock once and had never raised the issue again. Buck, on the other hand, called on us each year, and it finally got to the point where Dad, who admired persistence, would give him five minutes and talk in a friendly way.

  So when Williams and I decided to go ahead, it seemed reasonable to call Buck. He came rushing up instantly, brought a whole team in, and the thing got done. Morgan Stanley’s name appeared at the head of the list of underwriters, which was quite a coup because $200 million was the second-largest stock sale in Wall Street history. (The largest had been a $328 million offering by General Motors in 1955.)

  We hit the billion-dollar mark at the end of 1957, my first full year as chief executive. Selling a billion dollars worth of anything in those days was like supersonic flight—there weren’t many organizations that had achieved it. Only thirty-six American industrial companies—such as Procter & Gamble, Boeing, and Standard Oil—were bigger than IBM, and almost all of them depended on us for computers. It no longer took Watson optimism to say we were a significant factor in the U.S. economy. I allowed myself a small celebration. We had been wanting to upgrade our corporate airplanes, so we bought a Convair, a twin-engine fifty-passenger airliner that we fitted out with bunks and a meeting area. We used it to fly customers to sales demonstrations, not for junkets, but all the same it is heady stuff for a fellow of forty-three to buy a million-dollar airplane like that.

  I was terribly proud of the start I’d made. I didn’t like it when people compared me to my father, but I felt that if I could keep my record going for perhaps another ten years, I’d be able to count myself in the same league as Dad. So it came as a real slap in the face when one member of my family decided to hedge her bet on my making it. The second year after Dad died, my sister Jane sold a million dollars worth of IBM stock—about one third of her total holding. This was a real vote of no confidence, the first time any Watson had ever sold IBM stock, and it cut me so deeply that I went down to her house in Washington.

  “Of course it is your right to sell whenever you want,” I told her, “but why did you do it?” Jane was awfully surprised that I knew about the sale, because she didn’t understand corporations well enough to realize that a million-dollar sale is reported to the chief executive. But getting an answer out of her was exactly like having a discussion with the old man. Sometimes when you tried to pin him down, my father would give a totally ridiculous answer. If you said, “Dad, you picked up my bags and left me at the station!” his answer might be, “Oh, I thought you wanted to walk home!” That’s the kind of impossible response I got from Jane: “I didn’t think you’d be interested!”

  “How can you possibly think that? I’ve been running this business and you’ve been a beneficiary of it. But you’ve never given the slightest hint that you think this record is worth tipping a hat to.”

  “Oh, Tom. You know I think that.”

  “So then why are you selling your stock?”

  “Because I have to protect my family’s future.”

  In retrospect, I can see Jane was probably following the advice of some financial counselor who convinced her it was prudent to diversify her holdings. But at the time her selling of the stock really knocked me back, and it was the end of warm feelings between us for a number of years.

  Management is no science; it is much too human a process ever to qualify. Since we built such sophisticated business machines, people tended to think of IBM as a model of order and logic—a totally streamlined organization in which we developed plans rationally and carried them out with utter precision. I never thought for a minute that was really the case. While the 1950s was a boom time for such fields as organizational engineering and systems analysis, that was not the kind of leadership the people of IBM expected, or the job for which I’d been trained. Dad had taught me that a good businessman has to be an actor. You have to make a show of getting angry a lot more often than you really lose your temper; you have to look more worried than you really are when trying to stimulate someone to tackle a problem. Dad was a master at hamming it up in this way, and I patterned my actions after him whenever I got the chance.

  As IBM became huge and decentralized, the challenge was to find ways of maintaining personal contact with the people of the company, and to motivate other top executives to follow my example. One July an airliner carrying a number of IBM employees turned over taking off from Rochester, New York, in a thunderstorm. Seven people died in the crash,
including one IBMer, and eight or nine others from our group were hospitalized. I was at an IBM conference in Vermont when word of the accident came, and I checked quickly to make sure the chiefs of the divisions involved had gone to Rochester to help. But one executive was still in his Westchester office. I got him on the phone and said, “Are you going to go up there and go from bed to bed, or will I?”

  “Jeez, Tom, I never thought of it.” He’d been in the business a long time, and I asked how he could possibly have forgotten the example Dad had set after the Port Jervis train wreck in 1939—getting out of bed in the middle of the night and driving to the hospital to see the IBM families involved. “I’m giving you fair warning,” I said. “You’re supposed to run your division as if it were your own company, but if you’re not in Rochester by the end of today, I’m going myself.”

  “I’ll call you from the hospital,” he said, and he did about four hours later.

  Like Dad, I went to dozens of “family dinners” for IBM employees each year, and paying visits to local offices was standard procedure for me on any business trip. I kept a whole squad of secretaries busy doing things in the same way Dad would have—making sure every letter I received was acknowledged within forty-eight hours, sending flowers to the hospitalized wife of an employee, and making thousands of other small gestures of thoughtfulness. Dad often answered his own phone himself, and so did I whenever I could. If the caller was a customer it was a delightful surprise to reach Watson directly; and if the caller was an IBMer it set a powerful example for him to treat his own callers with equal consideration. Probably an efficiency expert would have condemned these practices as a gargantuan waste of time for a chief executive. But in a service-oriented business like ours, these seemingly minor details of courtesy and style were too important to let slide. If the head man stops taking a proprietary interest in them, pretty soon everybody else stops caring too.