Part IThe Story
A Radio in Boiling Water
The man the storytellers call "this Texan" walks into an office at the U.S. Army Signal Corps Laboratory at Fort Monmouth, New Jersey, carrying more than a salesman normally carries. The office belongs to Howard Moss. The visitor takes out his props one at a time: a beaker, a jug of water, a Bunsen burner, and a transistor radio. He pours the water and lights the burner. Then he waits, because water does not hurry, and a demonstration staged in front of a government buyer depends on the buyer having time to wonder what is coming. When the water reaches a rolling boil he switches the radio on and drops it into the beaker.
It keeps playing.
— Patrick E. Haggerty, to Howard Moss of the U.S. Army Signal Corps Laboratory, as recounted by the Chip History CenterMr. Moss, I want to talk to you about reliability.
Moss, the story goes, wrote him a check for a contract that same day.
A skeptic should say plainly that this account has been worn smooth by retelling. It survives in its fullest form as a few lines in the Chip History Center's hall of fame, which inducted Patrick Eugene Haggerty in 2016, the year it opened its doors to people who had "changed the business" rather than only those who had made chips. No date is attached to the meeting. Nobody records which radio it was, though it is tempting to picture one of the pocket sets his own company would put on the market for Christmas in 1954, and the sources do not support that picture. The props are too good, the line is too good, and the check arrives too quickly. The scene has the shape of a parable, and parables tend to lose their awkward parts in the telling.
The parts that remain, though, are the ones that held up under every retelling. The military needed transistors it could rely on. According to the same account, the established electronics companies, the radio-tube firms with their large plants and lineages going back before the war, were not supplying them. "They simply moved too slow." The man with the Bunsen burner was not one of the inventors of the device in the beaker. Bell Telephone Laboratories had invented the transistor. His company had bought the right to make it for $25,000. His contribution was to understand, more exactly than anyone else in the room, what the person holding the checkbook was afraid of.
He knew that fear well because he had spent the war on the buyer's side of the desk.
He was also not a Texan. He was born in Harvey, North Dakota, on March 17, 1914, which was St. Patrick's Day, and he was named Patrick. He studied in Milwaukee and learned management in a carton factory. He became an adult professionally inside a Navy procurement office in Washington. "Texas" reached his life by way of a trademark conflict, in a city where no electronics industry existed until he and a few men from the oil patch made one. The company he shaped for thirty-five years would make the first commercial silicon transistor, the first commercial transistor radio, the first integrated circuit, the handheld calculator, and an early single-chip microprocessor. At his death in 1980 the inventor's Nobel Prize was still twenty years off, and it went to someone else. Haggerty received medals named for other people and a museum named for himself and his wife.
He also received a reputation that his obituarists summed up in an unusually tender phrase: "a warm, kind, and gentle man, not without a touch of Irish temper when faced with obstacles to achievements or principles." The same company that operated under that gentleness built the guidance system for the first laser-guided bomb.
Every version of the story comes back to the radio in the beaker, still playing.
By the Numbers
The Haggerty Years at Texas Instruments
$25,000Price of the 1952 Western Electric patent license to make germanium transistors
$3M → $45MCompany sales, 1946 to 1956, when TI led the industry in transistor sales
~2 yearsTI's lead over other Bell licensees in commercial-quantity transistor production
Sept. 12, 1958Jack Kilby demonstrates the world's first working integrated circuit at TI
47 plants / 19 countriesTI's overseas manufacturing footprint during Haggerty's tenure
3 lbsWeight of the 1967 Cal-Tech, the handheld calculator prototype
$17.7BTexas Instruments revenue in 2025, nearly half a century after his death
The Telegrapher's Son
Harvey is a railroad town on the North Dakota prairie, and Michael Eugene Haggerty was a railroad telegrapher. The National Academy of Engineering's memorial tribute records that, along with his wife's name, Lillian Evenson, and very little else. A telegrapher's job was to translate. Electrical pulses came in over a wire from somewhere far beyond the horizon, the operator turned them into words, and the words moved freight, schedules, and news through country that otherwise had nothing to tie it together. Growing up in that house meant growing up with the idea that information arrives as current and that somebody's skill turns it into meaning.
The son went further than his father in this. Before he was grown he built a radio that won a prize, and he became, by the Texas State Historical Association's account, one of the first ham radio operators in North Dakota. That means a boy alone in a prairie house, listening through static for a voice from Minneapolis or Manitoba, and occasionally transmitting into the dark to see whether anyone answered. He also earned the rank of Eagle Scout. His memorialists later treated that achievement as the first instance of a lifelong method, and the line they chose to illustrate it reads more like a creed than a recollection:
"But we can only do it slowly, with persistence, with discipline, and, most of all, with wisdom."
We do not know what "it" referred to. The NAE tribute gives the sentence without its beginning, the way a sermon is remembered by its last line. It is a strange motto for a man whose company would become a byword for speed, the outfit that outran Bell Labs' other licensees by two years and announced a silicon transistor before the industry believed one could be built. The contradiction is only on the surface. Haggerty was fast where speed was cheap and patient where patience paid, and the skill lay in telling those two situations apart.
He arrived at Marquette University in Milwaukee in 1931 on a scholarship, in the deepest part of the Depression, and paid his way with a part-time job at the Badger Carton Company. The detail sounds trivial. Badger made boxes, and what a carton plant teaches is unglamorous and permanent: cost per unit, throughput, scrap, and the gap between what a design promises and what a production line delivers. He graduated summa cum laude in electrical engineering in 1936 and went to Badger full time as production manager. Within a year he was responsible for every engineering, manufacturing, and administrative function at the company, except sales.
The exception is the joke a novelist would be too embarrassed to use. The man who would later boil a radio in front of an Army buyer and sell transistors to IBM, the Air Force, and American teenagers was, at twenty-three, trusted with everything at a Milwaukee box company except selling the boxes.
On February 26, 1938, he married Beatrice E. Menne. They had five children and, by the time of his death, thirteen grandchildren. The Catholic thread in his life never broke and seldom showed in public: a Jesuit university in Milwaukee, then honorary degrees from the Catholic University of America and the University of Notre Dame, trusteeship and academic stewardship at the University of Dallas, and burial at Calvary Hill Cemetery. When Beatrice later set up the family foundation, she wrote its mission in language her husband's statements of corporate purpose never used: "to serve God by serving others."
He seems to have kept those two vocabularies apart, God in one and customers in the other, and the separation is part of what makes him difficult to see clearly. The prairie, the Jesuits, and the factory floor made up the early record. The war changed what he did with them.
The Buyer Changes Sides
In 1942 Haggerty went to Washington as an ensign in the U.S. Naval Reserve, assigned to the Bureau of Aeronautics. His first job there was to evaluate the manufacturers who supplied electronic equipment to the Navy. He was a grader. He sat on the receiving end of the American electronics industry's wartime output and wrote down who delivered, who slipped, whose equipment failed, and whose executives promised more than their plants could make. The work was patient and bureaucratic, and it was the best education in industrial reality available in the country. He went on to become responsible for all procurement and production of naval airborne electronic equipment and head of the Bureau's Electronics Production branch, rising to lieutenant. By the end of the war, every radio, radar, and electronic instrument that went aboard a Navy aircraft passed through an office he ran.
Consider what that meant. A man thirty years old knew the failure rates, delivery problems, and quality habits of nearly every serious electronics manufacturer in the United States. He knew which large firms were slow and why. He also knew, in a way no engineer at those firms could, what a military buyer actually lost sleep over: equipment that stopped working when it was hot, cold, wet, shaken, or far from a repair depot. When he later told Howard Moss that he wanted to talk about reliability, he was talking about the subject he had spent the war grading.
One of the suppliers he dealt with was a small Dallas company with an odd lineage, and its representative was J. Erik Jonsson.
Geophysical Service's ancestry ran through two physicists, John Clarence Karcher and Eugene McDermott, who had developed a seismographic process for oil exploration. In 1924 they started the Geophysical Research Corporation in Tulsa as a subsidiary of Amerada Petroleum. In 1930 they moved to Dallas and founded an independent firm, Geophysical Service, Incorporated, the first independent contractor to specialize in the reflection-seismograph method. The method is a form of listening. A crew sets off a charge at the surface, sound travels down through the earth, and instruments record the echoes that come back from buried rock layers. Oil companies paid GSI to listen to the ground and tell them what it said.
In 1939 the business was reorganized as Coronado Corporation, an oil company with GSI as a subsidiary. On December 6, 1941, the day before the Japanese attack on Pearl Harbor, McDermott and three other GSI employees, Jonsson, Cecil H. Green, and H. B. Peacock, bought the geophysical company. Nobody signing those papers on a Saturday in December could have known what the next morning would bring, and nobody could have known that the purchase would end up mattering less for oil than for war.
Jonsson had been hired to run a GSI laboratory in New Jersey. At the start of the war, according to the Texas State Historical Association, he saw something that, looking back, appears to be the founding insight of a great deal of American technology: the tools that listened for echoes from rock could also listen for submarines. A seismograph and a sonar both send energy into a medium and interpret what comes back. In 1942 GSI began building electronic systems for the military, starting with submarine detection gear and expanding into equipment for the Army, the Army Signal Corps, and the Navy.
So the meeting between Haggerty and Jonsson was a meeting between a buyer and a seller. The Texas State Historical Association states it bluntly: Haggerty had been "a purchaser of GSI's equipment for the United States Navy." The two men talked, the same account says, about the need after the war for "a good, small company to combine electrical and mechanical technology for innovative military and civilian products."
A good, small company. It is one of the few phrases from Haggerty's founding conversations to survive, and it is worth noticing because of what it leaves out. It says nothing about oil and nothing about Texas. It also says nothing about the transistor, which did not yet exist. The words that matter are military and civilian, both markets, from the beginning. And the phrase good sits in front of small the way a craftsman's adjective sits in front of the noun for a tool.
In November 1945, three months after the war ended, Haggerty moved his family to Dallas and became general manager of GSI's newly formed Laboratory and Manufacturing Division. His assignment covered research, engineering, and manufacturing for the whole company. The buyer had moved to the seller's side of the desk, and he brought his grading sheets with him.
An Accident of Trademark
For six years the Laboratory and Manufacturing Division grew inside a company whose identity belonged to something else. GSI was still a seismic firm, with crews in the field and instruments that listened to the ground, and L&M was the workshop attached to it that turned out anti-submarine equipment and oilfield gear for whoever would pay. Sales across the company in 1946 came to roughly $3 million.
Then the workshop began to outgrow the house. By 1951, carried by defense contracts, the L&M division was expanding faster than the geophysical business that had created it. In December 1951 the company reorganized around the fact that manufactured products had overtaken services in sales. The geophysical operation became a wholly owned subsidiary of a new parent, and the new parent needed a name.
The founders picked General Instruments. It was a bland, ambitious, East Coast sort of name, the kind of label a company uses when it wants to sound as if it might someday make anything. It was also unavailable, because a firm called General Instrument already existed. So the company became Texas Instruments, in the same year, by default.
This is the second piece of wry luck in the story, after the box company that kept Haggerty away from sales. The most famous geographic brand in American electronics was a fallback. The founders' first instinct was to sound general, and the trademark office pushed them into sounding particular. A company in Dallas, more than a thousand miles from Bell Labs, from the radio-tube giants, and from the Northeastern universities and capital markets that dominated American electronics, ended up named for the place that marked it as an outsider. "Texas Instruments" told everyone where the company was and therefore where it was not.
Haggerty became executive vice president and director of the new company. Jonsson was president. McDermott and then Green took turns as chairman. The arrangement suited the men involved: Haggerty was the operator, the builder of the manufacturing side, the hired general manager who had become indispensable. Jonsson, whom the company would later honor with a tribute of its own and the management profession with the Henry Laurence Gantt Medal in 1968, was the steadier corporate figure. McDermott and Green were scientists who had become owners. Haggerty was a factory man who had become, without anyone quite deciding it, the strategist.
In 1953 the company listed on the New York Stock Exchange. By then the business the name was originally meant to cover, seismic instruments and military electronics, was about to be overshadowed by a product the founders had not had in mind when they chose it.
Two dates in the corporate record are worth setting side by side. In 1951 the L&M division displaced its parent from inside. In 1988, after selling GSI and buying it back, Texas Instruments sold most of the old geophysical company to Halliburton, and GSI ceased to exist as a separate entity. The company that had listened to the ground for oilmen was absorbed into an oilfield services giant, and the instrument shop that had sprung from it had become one of the major semiconductor manufacturers in the world. In between lay most of Haggerty's working life, and the reversal came almost entirely from one wager he made in 1951.
Twenty-Five Thousand Dollars and a Crystal Grower
Bell Telephone Laboratories, the research arm of AT&T, had invented the transistor in the late 1940s. Under pressure to share it, Western Electric, AT&T's manufacturing arm, offered licenses to other companies. Texas Instruments paid $25,000 in early 1952 for the right to make germanium transistors.
The license was not the bold part. Many companies bought one. The bold part was Haggerty's reading of what it was for. According to the Texas State Historical Association, he believed that for TI to matter as an electronics company, it would need "to do more than assemble components." Assembling was what a good, small instrument shop did. It bought tubes, resistors, and wire from larger companies and built them into finished equipment, so its margins and fate depended on its suppliers. Haggerty wanted TI to become a supplier, to make the parts that everyone else would assemble. The license was the way in.
He saw something else as well, which the NAE memorial puts most sharply: "It was Pat Haggerty who recognized the way electronics would be revolutionized and proposed that the fledgling enterprise become a licensee." Proposed is the important word. He was an executive vice president making a case to owners who were geophysicists. A company that had earned its living listening to rock was being asked to bet on a three-terminal crystal device whose commercial value, in the words of IEEE Spectrum's history, "was still much in question."
A license, though, is a recipe, and recipes do not cook themselves. In 1952 Haggerty brought Gordon Teal to Dallas from Bell Labs.
Teal was a physical chemist whose expertise was the part of transistor making that most licensees underestimated: growing the semiconductor crystals. A transistor is only as good as the crystal it is cut from, and growing large, pure single crystals was difficult, slow, and close to an art. Teal had done it at Bell. Leaving Bell Labs, the most prestigious industrial laboratory in the country, for a geophysical company in Dallas was, by any reasonable professional measure, a step down. Teal took it. His first assignment was to direct TI's research laboratory.
Haggerty had not hired a general physicist or a manager. He had hired the man who had already solved the hardest step in the process, and in doing so he imported a piece of Bell Labs' tacit knowledge in the most reliable container available, a person. By the end of 1952 TI announced that it had grown to 2,000 employees and $17 million in sales. By the second half of 1953 the company was mass-producing germanium transistors under the direction of Mark Shepherd Jr., a production engineer who would serve as president from 1967 to 1976 and chairman from 1976 to 1988. In TI's leadership records, Shepherd's name follows Haggerty's in both columns.
Teal hired in turn. Early in 1953 Willis Adcock arrived, another physical chemist, and took charge of a small group aimed at a goal most of the industry considered premature: grown-junction silicon single-crystal transistors. Germanium was the established material. It was easier to work and better understood, and it was what Bell's licensees were all building. Silicon was harder to purify and harder to grow, but it could withstand the high operating temperatures that came with complex systems such as computers and, though nobody needed to say it aloud in Dallas in 1953, military hardware that had to work in a cockpit or a missile. Adcock later became TI's first Principal Fellow, the company's highest technical rank, an honor created in part to describe people like him.
The race that followed shows how industrial history is actually written. In January 1954, at Bell Labs, Morris Tanenbaum built the first workable silicon transistor. The result was reported that spring at the Institute of Radio Engineers' off-the-record conference on solid-state devices and appeared later in the Journal of Applied Physics. Bell got there first and told only a closed room.
Working independently, Teal's people produced what would be recognized as the first commercial silicon transistor and tested it on April 14, 1954. On May 10, 1954, at the IRE National Conference on Airborne Electronics in Dayton, Ohio, Teal presented a paper titled, with a chemist's literalness, "Some Recent Developments in Silicon and Germanium Materials and Devices." TI announced production of two types of silicon transistors. In the Texas State Historical Association's description, the company "against conventional wisdom, surprised the industry."
Notice the venue. It was a conference on airborne electronics, which was the field Haggerty had overseen for the Navy and the market he understood better than anyone in his own company. It was also on the record. Bell had the priority, and TI had the announcement, the product, and the customers in the audience. For decades afterward the common phrase was that TI made the first silicon transistor, and that is approximately true in the way commerce defines first: the first one available for purchase.
The NAE memorial gives the outcome its proper scale. Haggerty "selected, developed, and led the group of scientists and engineers whose work enabled Texas Instruments to manufacture germanium and silicon transistors, in commercial quantities and approximately two years ahead of other licensees." Every licensee paid the same $25,000. The two-year lead was what Haggerty bought with Teal's crystals, Adcock's silicon, Shepherd's production lines, and his own insistence that TI be a maker of parts.
Still, the parts were expensive, and expensive parts needed someone who wanted them.
Christmas, 1954
Making transistors in commercial quantities was not enough to bring prices down, and the early prices were high. Haggerty's diagnosis was again a manufacturer's diagnosis. The way to lower cost was volume, and the way to get volume was a market large enough to soak it up. That market did not exist. Radios, televisions, and hearing aids ran on vacuum tubes, which were bulky, power-hungry, and familiar. Nobody was asking for transistors in quantity because nobody had a product that required them.
So Haggerty created one.
In May 1954 Texas Instruments designed and built a prototype of a transistorized pocket radio. TI did not intend to become a radio company. It partnered with Industrial Development Engineering Associates of Indianapolis, a small firm with a big name, and IDEA sold the set under the Regency brand. The Regency TR-1 went on sale in October 1954, timed, in the Texas State Historical Association's phrase, "just in time for Christmas." It was entirely solid-state and small enough to carry in a pocket.
It ran on germanium. The company that had just stunned the industry with silicon built its consumer showcase on the older material, because silicon transistors cost much more at the time. The decision deserves notice because it shows the two halves of Haggerty's thinking operating together. Silicon was the strategic direction, the material for computers, military systems, and the future. Germanium was what a Christmas gift could afford. He did not let the better technology stand in the way of the bill of materials. He let each material do the job it could do in that year.
The Regency was a demand-generation device presented as a radio. Wikipedia's history of the company states it directly: the radio "was an effort by Haggerty to increase market demand for transistors." Every set sold was a transistor sold, and every transistor sold pushed the next one down the cost curve. Every teenager carrying a Regency was a demonstration, to the people who designed radios, televisions, and computers, that the vacuum tube could be replaced in a mass-market product by a component made in Dallas.
The radio pulled the company into the industry it actually wanted. TI soon became a primary transistor supplier to IBM, which was growing quickly. Sales reached $45 million in 1956, fifteen times the 1946 figure, and TI led the industry in transistor sales. Harvard Business School, compiling its list of the great American business leaders of the twentieth century, credited Haggerty with leading TI into transistor manufacture as "the first to make them cheaply enough to be commercially viable."
Cheaply enough is the phrase that sums up this part of his career. Bell Labs invented the transistor, TI made it affordable, and Haggerty saw sooner than his competitors that the second accomplishment would be worth more than the first to everyone who did not own AT&T.
There is a rhyme here, and the story has several like it. Three decades earlier a boy in North Dakota had built a radio and won a prize for it. Now the man had built a radio and won a market. In between, at a desk in Washington, he had overseen the production of every radio that flew on a Navy aircraft. The device kept recurring in his life, each time a little smaller and each time carrying a little more of his ambition.
The Hired Wanderers
Haggerty believed, in the Texas State Historical Association's words, "that electronics could benefit all parts of society." That reads like a slogan until you look at what he did with the belief, which was to pay people to find out where it led. "Discussions led to the hiring of research engineers to explore the possibilities," the same account says. The verb is explore, and the mandate was deliberately vague. A company still dependent on defense contracts and transistor margins hired engineers without asking them to deliver a specific product.
One of them was Jack Kilby.
In July 1958 Kilby recorded his first ideas about building an entire circuit, not just a transistor, from a single piece of semiconductor material. Resistors, capacitors, and transistors would all be carved from the same block instead of made separately and wired together by hand. On September 12, 1958, he demonstrated the first working integrated circuit. It was made of germanium, the humble material once again. About six months later Robert Noyce of Fairchild Semiconductor in California independently developed an integrated circuit in silicon with integrated interconnections, which turned out to be a far more practical way to manufacture and wire the components.
What followed is the classic story of American technology in the second half of the twentieth century: two inventors, two companies, and a ten-year legal fight that ended with both sides signing joint patent agreements. Kilby is credited with the idea of integrating components on one chip and Noyce with the practical method of connecting them. In 1968 Noyce co-founded Intel. In 1969 Kilby received the National Medal of Science, in 1982 he entered the National Inventors Hall of Fame, and in 2000 he received the Nobel Prize in Physics. In 2008 TI named its development laboratory Kilby Labs.
The pattern in that list is easy to miss. Haggerty created the conditions for the invention and did not share in the credit, which is how such arrangements usually work. In the official history the integrated circuit has two fathers, Kilby and Noyce. The man who decided a Dallas company should pay engineers to wander does not appear among them.
Haggerty seems to have expected this. "Always a team man as well as a leader," the NAE tribute says, choosing team over any word that would suggest ego. The integrated circuit was at once a product of his intent and a product of things he could not foresee. He hired people to explore because he could not name what they would find, and they found the thing that would make him matter.
Meanwhile the silicon wager paid off in a series of practical contracts. TI's work on silicon transistors and integrated circuits rated for military temperature ranges won the company a contract in 1961 for the first integrated-circuit-based computer for the U.S. Air Force, a demonstration machine TI promoted as the Molecular Electronic Computer. In 1962 TI won the contract to supply integrated circuits for the Minuteman missile. In the 1960s TI developed the 7400 series of transistor-transistor logic chips, standardized components that made integrated circuits the common material of computer logic. Silicon, chosen in 1953 partly because it could survive heat, was now going inside intercontinental ballistic missiles.
In 1959 the company made what the Texas State Historical Association called "its only major merger," with the Metals and Controls Corporation of Attleboro, Massachusetts. The Attleboro business became TI's Sensors & Controls division. In 2006 Bain Capital bought it for $3 billion in cash, and it became Sensata Technologies. Even Haggerty's rare acquisitions had long afterlives.
In 1958 he became president of Texas Instruments, and by the end of the next decade he was chairman. Harvard dates his leadership era from 1958 to 1976. The men who had once been the company's officers took ceremonial roles, and the operator ran the company in name as well as in practice.
Listening Devices
It is uncomfortable to put the warm, gentle, devout Haggerty next to the full list of what his company built, and the discomfort is worth keeping.
In 1956 TI began research in infrared technology. That work produced line-scanner contracts, and in 1963 TI invented the first forward-looking infrared system, FLIR, which let a pilot see the heat signature of a landscape at night. Production began in 1966. In 1961 TI won the guidance and control contract for the AGM-45 Shrike, an anti-radiation missile designed to home in on enemy radar emissions. In 1964 it began developing the first laser guidance system for precision-guided munitions, which led to the Paveway series of laser-guided bombs, starting with the BOLT-117. In 1966 the company began MERA, Molecular Electronics for Radar Applications, the first solid-state radar. In 1968 it built data systems for the Mariner planetary probes. In 1969 it won the seeker contract for the Harpoon missile.
Each of these systems is, at bottom, a way of listening: detecting heat, detecting a radar beam, detecting reflected laser light and steering toward it. The company that had started by listening for echoes from rock had become a company that listened for the enemy.
The most literal case came in 1965. Under a top-secret government contract, GSI, the old seismic subsidiary that had become an asset of the semiconductor company, took part in Vela Uniform, a program to detect Soviet underground nuclear tests and verify compliance with the Partial Nuclear Test Ban Treaty. The same methods developed for finding oil were now listening through the earth's crust for bombs on the other side of the world. The listening technique had gone from the oil field to the submarine to the warhead, and in every case the information came back as a faint electrical signal that needed interpreting, like a telegraph pulse arriving at a prairie depot.
Defense paid the bills and also, as Haggerty had foreseen in his "military and civilian" conversation with Jonsson, set the standards. Military buyers demanded reliability, temperature range, and silicon, and the discipline of meeting their specifications became the company's manufacturing culture. The consumer markets demanded something else: price, volume, and the willingness to stage a Christmas launch.
He held the two together in a single sentence, the founding statement of purpose of Texas Instruments, which contains no slogan and no mention of either weapons or toys:
— Patrick E. Haggerty, Texas Instruments Statement of PurposeTexas Instruments exists to create, make, and market useful products and services to satisfy the needs of its customers throughout the world.
The order of the verbs is create, make, market. That is invention, then manufacturing, then sales, the sequence of his own career compressed into three words: the ham radio boy, the carton plant manager, the man with the Bunsen burner. Useful is the only adjective, and it does a great deal of work. It is broad enough to cover a pocket radio, a classroom calculator, and a laser-guided bomb, and it does not ask which one. Whether that breadth was a moral position or a refusal to take one, the record does not settle. The memorial tributes do not take up the question, and Haggerty, as far as his surviving words show, did not either.
What the record does show is a man who took on the profession as a second job. In 1962 he served as president of the Institute of Radio Engineers. The name of the organization recalls the boy in North Dakota, and the IRE was where Teal had made his announcement in Dayton. Haggerty co-chaired the committee that merged the IRE with the American Institute of Electrical Engineers to form the Institute of Electrical and Electronics Engineers. The new IEEE gave him its Founders Medal in 1968, with a citation that is more bureaucratic than lyrical and therefore probably accurate:
— IEEE Founders Medal citation, 1968For outstanding contributions to the leadership of the electrical and electronics engineering profession, with special reference to the development of the worldwide semiconductor industry and service to the profession through his contributions leading to the creation of the IEEE.
Worldwide was meant literally. Harvard counts 47 manufacturing plants in 19 countries during his tenure, and describes the company as implementing "Japanese-style management techniques to form the 'TI culture.'" The IRI Medal followed in 1969, the John Fritz Medal in 1971, and in 1975 the Henry Laurence Gantt Medal for management, which Peter Drucker had received in 1959, Jonsson in 1968, and David Packard in 1969. The engineering profession honored him as an engineer, and the management profession honored him as a manager. He belonged fully to neither group, which is likely why both wanted to claim him.
Productivity per Man-Hour
In 1965 Haggerty, as chief executive, launched a research and development program with a goal that sounded like a toy: a calculator small enough to hold in one hand. He put a team on it that included Jack Kilby. By 1967 Kilby and two colleagues had built the Cal-Tech, a battery-powered prototype that weighed three pounds, measured 4.25 by 6.15 by 1.75 inches, did basic arithmetic, and showed six-digit numbers. TI brought a product to market in 1971. The SR-10, named for the slide rule it was meant to replace, followed in 1973, and the SR-50 scientific calculator in 1974. Both had red LED digits.
The Regency logic had come back almost unchanged. In 1954 a pocket radio created demand for transistors. In the late 1960s a pocket calculator would create demand for the integrated circuits that only a few companies could make in volume. The processor developed for the Cal-Tech would, in Wikipedia's account, give rise to the great majority of TI's later revenue. Haggerty was again building the device that would sell his components to millions of people who had never heard the word semiconductor.
The calculator line led toward the microprocessor. TI introduced a single-chip microcontroller in 1970 and a single-chip microcomputer in 1971, and on September 4, 1973, it was granted the first patent on a single-chip microprocessor, credited to Gary Boone. Intel, Noyce's new company, is usually credited alongside TI for the nearly simultaneous invention, which makes the second time in fifteen years that Dallas and Santa Clara reached a fundamental device at about the same moment. The microprocessor patent outlived Haggerty and survived a challenge as well. A competing patent awarded to Gilbert Hyatt in 1990 was overturned in TI's favor on June 19, 1996.
During those same years Haggerty wrote something unusual for an electronics chairman. In two papers on productivity in American education, "Education, Work, and Productivity" and "R & D and Productivity in Education," he set out an argument that must have sounded very cold to some teachers who heard it. Productivity per man-hour in the private economy had been rising, he observed, while productivity per man-hour in education had fallen, because educators kept pressing for lower student-teacher ratios. He suggested that technology, video cassettes and instructional television networks, could carry teaching beyond the classroom, improve results, and raise the number of students per teacher. In the second paper he called for moving from a teacher-oriented system to a student-oriented one, backed by large-scale research and development on student-centered learning.
This was the carton plant manager thinking about schools. It was also, quietly, the strategist thinking about a market. In the decade after his death, TI worked with the National Council of Teachers of Mathematics to develop a calculator that could become the educational standard. Connecticut required graphing calculators on state exams in 1986. Chicago's public schools gave a free calculator to every student from fourth grade on, beginning in 1988. The College Board required calculators on Advanced Placement tests in 1993 and allowed them on the SAT in 1994. The TI-81, released in 1990, made TI the leader in graphing calculators, and generations of American teenagers would sit exams with a Texas Instruments device on the desk. Estimates put the production cost of a TI-84 Plus at $15 to $20, with a margin likely above 50 percent.
None of that happened while Haggerty was alive, and assigning it to him would be a form of hagiography. Still, the papers exist. The chairman who argued that students should be at the center and that technology could make each teacher more productive presided over the company that, a decade later, made itself indispensable to the American math classroom. It is hard to say whether the later monopoly fulfilled his argument or debased it. Probably it did some of each.
One more device belongs here, because it came out two years before he died. In 1976, while studying uses for the bubble memory then being developed, TI engineers narrowed their focus to speech. The result was the TMC0280, a single-chip linear predictive coding speech synthesizer and the first time a single piece of silicon had electronically reproduced the human voice. It shipped in a child's educational toy, the Speak & Spell, introduced at the Summer Consumer Electronics Show in June 1978. A child would press keys, and a chip would pronounce a word in a flat voice that no throat had ever produced, then wait for the child to spell it.
The telegrapher's son, whose father had turned electrical pulses into words at a railroad depot, had lived to see his company build a machine that turned electrical pulses into speech, and had sold it as a spelling lesson.
The Last Lecture
On February 13, 1980, Haggerty, then honorary chairman of Texas Instruments, gave a lecture to NASA in Houston. He was sixty-five. His subject was the history and philosophy behind the strategic management systems he had built at TI, the planning machinery through which a seismic contractor had remade itself several times and which had, in the words of the journal that later published the talk, "shaped policy within this highly innovative organization." He had been a founding member of the editorial board of the Strategic Management Journal. Practitioners of his generation did not usually write for academics. He did, and the academics treated him as one of their own.
Eight months later he was dead. Patrick Haggerty died on October 1, 1980, at Presbyterian Hospital in Dallas after a short illness. He was sixty-six. He had been on Texas Instruments' board since the company's founding in 1951, and his career there had lasted thirty-five years, beginning in a month when the war was barely over and ending at the start of the personal computer era.
In April 1981 the Strategic Management Journal printed the NASA lecture under the title "The Corporation and Innovation," pages 97 through 118 of its second volume, "as a mark of our respect and in memoriam." The editors called his death "untimely." The adjective makes more sense alongside the other founders' careers. Jonsson's long tenure as chairman ran to 1969. Haggerty, the youngest of the four men later counted as founders and the one hired rather than present at the 1941 purchase, outlived his own chairmanship by only four years.
What happened next at Texas Instruments reads like a warning written in his own terms. In 1979 the company entered the home computer market with the TI-99/4. Its successor, the TI-99/4A, arrived in 1981. By late 1982 TI dominated the U.S. home computer market and was shipping 5,000 machines a day from a factory in Lubbock. In late 1983, after a brutal price war fought mainly against Commodore, it discontinued the line. The Texas Monthly headline was "Death of a Computer." The digital watches had already gone the same way: early quality, poor battery life, then styling problems, too many models, and price points that could not compete. The founders of Compaq, which became an early leader in IBM-compatible PCs, all came from TI.
The Regency strategy, driving volume to lower cost and creating demand with a consumer showcase, worked when the consumer product existed to sell components. It failed when the consumer product became the business in its own right, competing on price against companies that cared about nothing else. Haggerty's method had always kept the radio as the means and the transistor as the goal. Without him, the company seems to have confused the two for a while.
It recovered. It sold the defense business to Raytheon in 1997 for $2.95 billion, sold Sensors & Controls in 2006, and bought National Semiconductor in 2011 for $6.5 billion, which made it the world's largest maker of analog components. In 2025 it reported $17.7 billion in revenue. The current company describes itself as a maker of chips for other people's products, a supplier rather than an assembler, which is the role Haggerty argued for when he proposed the $25,000 license.
His money went where his convictions had always gone, toward Milwaukee and toward schooling. Near the end of his life he chaired the board of trustees of Rockefeller University and helped lead the academic growth of the University of Dallas. He served as a director of the A. H. Belo Corporation, publisher of the Dallas Morning News. He and Beatrice left part of their estate to found the Patrick and Beatrice Haggerty Museum of Art at Marquette, the Jesuit university that had taken in a scholarship student from Harvey in 1931. Beatrice created the Patrick and Beatrice Haggerty Foundation to carry on their giving and involve the family. Between 1983 and 2016 it made more than 500 grants totaling more than $19 million, mostly in Texas and Wisconsin, the two states that bracketed his adult life.
He was buried at Calvary Hill Cemetery in Dallas. The company that bears the state's name is still there, along with the laboratory named after the engineer he hired to explore, the graphing calculators in the backpacks of teenagers who will never learn his name, and, nearly as old as all of it, a story that may or may not have happened as the industry tells it.
The story has a government office in New Jersey, a Signal Corps buyer who has not yet decided, a jug of water emptied into a beaker, and a Bunsen burner hissing on a desk. At the bottom of the beaker, in water at a full boil, there is a small radio, still playing.
Part IIThe Playbook
Haggerty did not invent the transistor, the integrated circuit, or the microprocessor, and he never claimed to. What he built was the machinery around invention: the license, the hire, the demand, the factory, the customer, the profession, and the sentence that held them together. The principles below are drawn from that machinery. They concern how a small company in an unlikely place repeatedly beat larger incumbents to commercial scale with technology those incumbents owned or understood first. Several have a darker side, and those are marked.
The principles
- 1.Sell the fear you once had as a buyer.
- 2.Buy the recipe, then outrun the cook.
- 3.Hire the person who has already solved the hardest step.
- 4.Build the product that creates demand for your real product.
- 5.Be dogmatic about direction, pragmatic about materials.
- 6.Announce on the record what others say off it.
- 7.Let the faster-growing division eat the parent.
- 8.Pay people to wander, and accept that they will get the credit.
- 9.Serve two masters whose demands make each other stronger.
- 10.Treat your profession as infrastructure.
- 11.Write the purpose in one sentence and the strategy down in full.
T
The Haggerty Sequence
From Navy buyer to semiconductor strategist: the decisions that compounded.
1942
Joins the Navy's Bureau of Aeronautics; evaluates electronics suppliers, later runs all airborne electronics procurement and production.
1945
Joins GSI in Dallas in November as general manager of the new Laboratory and Manufacturing Division.
1951
L&M outgrows its parent; the company becomes Texas Instruments after "General Instruments" proves unavailable.
1952
$25,000 Western Electric transistor license; hires Gordon Teal from Bell Labs.
1954
Commercial silicon transistor announced in Dayton (May 10); Regency TR-1 on sale in October.
1958
Becomes president; Jack Kilby demonstrates the integrated circuit (Sept. 12).
1962
President of the Institute of Radio Engineers; ICs for the Minuteman missile.
1965
Launches the handheld calculator program; Cal-Tech prototype follows in 1967.
1973
TI granted the first single-chip microprocessor patent.
1980
Delivers NASA lecture on TI's strategic management (Feb. 13); dies Oct. 1.
Principle 1
Sell the fear you once had as a buyer
Haggerty's wartime job was to grade the American electronics industry from the buyer's side: who delivered, who slipped, whose equipment failed in service. When he became a seller, he had a map of the customer's anxieties that no competing executive had. The boiling-water demonstration, whatever its exact history, did not pitch the transistor's elegance or its physics. It pitched the one thing a Signal Corps buyer had learned to dread, equipment that quit, and it did so with props the buyer could not argue with.
Most founders reason from the product outward. Haggerty reasoned from the purchase order back. The question was never "what does this device do?" but "what does the person approving the spend fear losing, and how do I make that fear disappear in front of them?" The transistor's actual advantage (no fragile tubes, lower power, smaller size) became persuasive only when translated into the buyer's terms.
The deeper lesson concerns career sequencing. Time spent as a customer, whether in procurement, as an operator using the tools, or on the receiving end of vendor promises, is not a detour from building. It is reconnaissance, and very few founders have done it.
Tactic: Before your next major sale, write down the three things your buyer is personally blamed for when they go wrong, and design your demonstration to make one of them visibly impossible.
Principle 2
Buy the recipe, then outrun the cook
Every Bell licensee paid the same $25,000 for the same transistor patents. The license conferred no advantage. Execution did: TI reached commercial-quantity production of both germanium and silicon transistors roughly two years ahead of the other licensees. Haggerty understood that in a technology made available to everyone, the race shifts at once from who knows how to who can make it at scale first.
This reframes the usual worry about "not owning the core IP." TI never owned the transistor. It became the transistor's most effective manufacturer, and in a fast-moving field a manufacturing lead compounds. Volume lowers cost, lower cost wins customers such as IBM, and those customers fund the next process improvement. Bell kept the priority and TI captured the market.
The license was also cheap relative to what it unlocked. A $25,000 entry fee into the most important industry of the century is the kind of asymmetric bet a small company should grab while larger companies are still debating whether the technology has commercial value.
Tactic: When a foundational technology is licensed or open-sourced broadly, stop competing on access and compete entirely on how quickly you can produce it reliably at volume.
Principle 3
Hire the person who has already solved the hardest step
A license is a recipe, and a recipe leaves out the hands. The hardest part of early transistor manufacturing was growing pure single crystals, and Gordon Teal had done it at Bell Labs. Haggerty did not hire a general scientist and hope. He found the specific bottleneck and recruited the specific person who had already overcome it, then put him in charge of research. Teal then hired Willis Adcock, another physical chemist, to go after the next bottleneck, silicon.
This is tacit-knowledge transfer done the only reliable way, inside a person. Documents travel easily and judgment travels with people. Teal's move from Bell to a geophysical company in Dallas was, by any status measure, a step down, which means Haggerty must have offered something prestige could not: authority over the problem.
Notice also the chain. Haggerty hired Teal, Teal hired Adcock, and the company's first Principal Fellow came out of that second hire. The highest-leverage recruiting move is the one that brings in someone who will make the next great hire.
Tactic: Identify the single hardest step in your production process, find the person anywhere who has already solved it, and offer them ownership of that problem rather than a title.
Principle 4
Build the product that creates demand for your real product
The Regency TR-1 was not a radio strategy. It was a transistor strategy. Transistors cost too much because volumes were too low, and volumes were too low because no mass-market product required them. Haggerty broke the cycle by designing the end product himself, partnering with a small Indianapolis firm to sell it, and timing it for Christmas 1954. Every radio sold pulled the cost curve down and showed every equipment designer in America that tubes could be replaced.
He did it a second time with the handheld calculator, launched as an R&D program in 1965, prototyped in 1967, and on the market by 1971. The calculator pulled integrated circuits into consumer volume just as the Regency had pulled transistors.
The danger appears in the sequel. When the showcase product turns into a business of its own, the logic changes. The Regency worked because TI did not try to become a radio company. The home computer wars of the early 1980s, fought after Haggerty's death, showed what happens when the showcase becomes the battleground and the company ends up in a price war against firms that do nothing else.
R
Showcase Products: When the Strategy Worked and When It Didn't
The demand-creation logic depends on keeping the consumer product subordinate to the component.
| Product | Real goal | Outcome |
|---|---|---|
| Regency TR-1 radio (1954) | Transistor volume and visibility | Scale TI leads transistor sales by 1956 |
| Cal-Tech / handheld calculator (1967–71) | Integrated-circuit volume | Scale Processor line drives later revenue |
| Speak & Spell (1978) | Showcase for single-chip speech synthesis | Try Iconic product; speech chips later sold off in 2001 |
| LED/LCD watches (1970s) | Consumer business in its own right | Risk Battery life, styling, model sprawl, price |
| TI-99/4A home computer (1981–83) | Consumer business in its own right | Risk Discontinued after price war with Commodore |
Tactic: If customers won't buy your component because no product requires it, build the minimum end product that makes the need visible, and partner to sell it so you never forget which business you are actually in.
Principle 5
Be dogmatic about direction, pragmatic about materials
In 1953 TI committed to silicon against the industry consensus, because silicon could survive the heat of computers and military systems. In 1954, the same year it announced the first commercial silicon transistor, it built its consumer showcase on germanium, because silicon was still too expensive. Kilby's first integrated circuit, in 1958, was germanium too.
This is not inconsistency. It reflects an understanding that a strategic direction and a bill of materials operate on different timescales. Direction is a multi-year bet on where the high-value markets will go. Materials are this quarter's arithmetic. Companies that merge the two either refuse to ship until the ideal technology is affordable, which kills momentum, or let today's cheap option set their long-term direction, which kills the future.
Haggerty kept both: silicon for the Air Force computer of 1961 and the Minuteman ICs of 1962, germanium for the Christmas radio of 1954. Each material did the job it could do in that year.
Tactic: Separate your roadmap into "the bet" (what you're building toward over five years) and "the bill" (what you ship this year), and never let either one veto the other.
Principle 6
Announce on the record what others say off it
Bell Labs built a working silicon transistor in January 1954 and reported it at an off-the-record conference. TI tested its own on April 14, 1954, and announced it on May 10 at a public IRE conference on airborne electronics in Dayton, with a production product behind the announcement. Decades of shorthand gave TI the "first."
The venue mattered as much as the timing. Airborne electronics was the market Haggerty knew from the inside, and that audience included the buyers who most needed heat-resistant silicon. A technical announcement made in front of customers works as a sales event as well as a scientific one.
Priority in the lab is a fact for historians. Priority in the market belongs to whoever stands up in public with something for sale. Incumbents with large research organizations often treat their discoveries as internal assets to be released when they are ready. An insurgent can take the public moment.
Tactic: When you reach a technical milestone, announce it on the record, with a purchasable product, at the venue where your most demanding customers are already in the room.
Principle 7
Let the faster-growing division eat the parent
By 1951 the Laboratory and Manufacturing Division Haggerty ran was growing faster than GSI's founding geophysical business. The owners did something most companies cannot. They reorganized the whole enterprise around the newer unit, renamed it, and made the original business a subsidiary. The seismic company that had paid for everything became an asset of the instrument company it had spawned, and in 1988 it was sold to Halliburton.
The decision was made by McDermott, Green, and Jonsson, the owners whose identity was tied to the old business. Haggerty's part was to grow L&M so persuasively that the reorganization became obvious. That is the operator's form of this principle: you do not argue for the pivot, you produce the growth curve that makes the argument for you.
Even the accident of the name supported the change. "General Instruments" was taken, and "Texas Instruments" gave the new company an identity that had nothing to do with oil.
Tactic: Track growth rates by business unit, not just size, and when a younger unit consistently outgrows the core, start designing the corporate structure that puts it on top.
Principle 8
Pay people to wander, and accept that they will get the credit
Haggerty's belief that electronics could serve "all parts of society" turned into a hiring practice: research engineers brought in "to explore the possibilities," with no specific product assigned. One of them, Jack Kilby, demonstrated the integrated circuit on September 12, 1958. Kilby received the National Medal of Science, a place in the Inventors Hall of Fame, the Nobel Prize, and a laboratory named after him. Haggerty received none of the inventor's honors and seems not to have wanted them.
This is the hardest kind of R&D investment for an executive to make, because the outcome cannot be named in advance, the payoff is uncertain, and success goes to someone else. It is also the only reliable way to produce discoveries that a roadmap could not have anticipated. Haggerty then aimed those discoveries at products: Kilby went on to the calculator team, and the calculator became a business.
The memorial phrase, "always a team man as well as a leader," describes the temperament this requires. A leader who needs to be the inventor will either crowd out inventors or attach to their work in ways that drive them off.
Tactic: Fund a small team with an open mandate in your core technology, judge it on what it surfaces over years rather than quarters, and make it publicly clear that discoveries belong to the discoverers.
Principle 9
Serve two masters whose demands make each other stronger
From his first conversation with Jonsson about "military and civilian products," Haggerty built TI to sell to two very different customers. The military demanded reliability, extreme temperature ranges, and silicon, and paid for development: the Shrike, Paveway, FLIR, Minuteman, Mariner. Consumers demanded low prices and volume: the Regency, the calculator. Each market made TI better at serving the other. Military standards made the components dependable, and consumer volume made them cheap.
The arrangement has a moral cost that the record does not resolve. The same company, operating under the same one-sentence statement of purpose, built children's spelling toys and laser-guided bombs. Haggerty's "useful products" covered both without distinction. Readers should take the strategic insight and keep the question open. A two-market structure is very effective, and it also spares a leader from having to say what the company is for.
The structure also needs active management. Once Haggerty was gone, the consumer side expanded into businesses (watches, home computers) that no longer helped the component side, and TI eventually sold the defense business to Raytheon for $2.95 billion in 1997.
Tactic: Look for a second customer segment whose requirements would discipline your weakest operating habit, whether quality, cost, or speed, and pursue it as a capability investment as well as a revenue line.
Principle 10
Treat your profession as infrastructure
While running a fast-growing company, Haggerty served as president of the Institute of Radio Engineers in 1962 and co-chaired the committee that merged it with the AIEE to form the IEEE. He sat on the founding editorial board of the Strategic Management Journal, chaired Rockefeller University's trustees, and wrote papers on productivity in education. Each looks like a distraction from the company, and each strengthened the environment the company depended on.
A semiconductor industry needs a professional body that sets norms, a talent pipeline, a shared technical vocabulary, and venues for announcements, and TI's silicon transistor debuted at an IRE conference. An executive who helps build those institutions shapes the field his company competes in. The IEEE's Founders Medal citation credits him with both "the development of the worldwide semiconductor industry" and "service to the profession," treating them as one contribution.
For an outsider company in Dallas, a long way from the established centers of electronics, this mattered more than usual. Professional leadership gave a Texas firm standing that its geography did not.
Tactic: Take on one real leadership role in the professional or standards body of your industry, and treat it as a strategic asset, not as philanthropy.
Principle 11
Write the purpose in one sentence and the strategy down in full
TI's statement of purpose fits on one line: "to create, make, and market useful products and services to satisfy the needs of its customers throughout the world." The verbs follow the sequence of the business (invention, manufacturing, sales), and the single adjective, useful, is the standard for every product. A sentence that short can be remembered, and so it can be used.
Haggerty also built, and late in life explained in detail, a system of strategic management that the Strategic Management Journal credited with "shaping policy within this highly innovative organization." His last major lecture, at NASA on February 13, 1980, was devoted to it. The pairing is the point: a purpose short enough to remember, and a strategy thorough enough to run a company of 47 plants in 19 countries.
Most companies have one without the other, either a moving mission statement with no operating system beneath it or an elaborate planning process with no sentence that tells people why it exists. Haggerty insisted on both, and he wrote both down himself.
Tactic: Reduce your company's purpose to one sentence built on the verbs of what you actually do, then document the strategic process that turns that sentence into decisions, in enough detail that you could lecture on it.
H
Conventional Wisdom vs. the Haggerty Approach
Where his choices departed from the incumbents of the 1950s electronics industry.
| Conventional wisdom | Haggerty's approach |
|---|---|
| Assemble equipment from others' components. | Become the component supplier; "do more than assemble." |
| Wait for demand to justify production. | Build the product that creates demand (Regency, calculator). |
| Germanium is the proven material. | Bet on silicon for the future; ship germanium when it's cheaper. |
| Report breakthroughs to peers, quietly. | Announce on the record, with product, before customers. |
| Protect the founding business. | Reorganize around the faster-growing division. |
| R&D on a defined roadmap. | Hire engineers "to explore the possibilities." |
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— Patrick E. Haggerty, as quoted in the National Academy of Engineering's Memorial TributesBut we can only do it slowly, with persistence, with discipline, and, most of all, with wisdom.
— National Academy of Engineering, Memorial Tributes, Volume 2It was Pat Haggerty who recognized the way electronics would be revolutionized and proposed that the fledgling enterprise become a licensee of Bell Telephone Laboratories for manufacture of transistors. Always a team man as well as a leader, he selected, developed, and led the group of scientists and engineers whose work enabled Texas Instruments to manufacture germanium and silicon transistors, in commercial quantities and approximately two years ahead of other licensees.
Maxims
- Sell from the buyer's chair. The person who has bought the product knows what to fear about it, and that fear makes the most persuasive demonstration.
- The license is the entry fee, not the moat. When everyone can buy the recipe, the advantage goes to whoever reaches volume first.
- Hire the solved problem. Find the bottleneck, then find the person who has already overcome it somewhere else.
- Make the market your component needs. If no product requires your part, build one and make sure you remember which business it serves.
- Bet on silicon, ship germanium. Long-term direction and this year's materials operate on different clocks, and both have to be respected.
- First is a public act. Priority in the lab belongs to historians, and priority in the market belongs to whoever announces with something for sale.
- Let the workshop take over the house. When a newer unit outgrows the core, rebuild the company around it.
- Pay for wandering, and give away the credit. Discoveries no roadmap predicted come from people who were never handed one.
- Two masters can make each other stronger. A demanding customer and a cheap customer, served together, can produce a product that is both reliable and inexpensive, but know what you are being made into.
- One sentence for the purpose, a full system for the strategy. A company needs a mission people can remember and a process that turns it into decisions.
