#419: Kelly Johnson: Skunk Works
What this covers
This episode finds David working through the writing of Kelly Johnson, the Lockheed engineer who ran the Skunk Works, and pulling out the working rules behind its aircraft. The stories are concrete. The first shop was built from engine boxes and a tent. The XP-80 was delivered on day 143 against a 180-day schedule, with about 120 people and 23 engineers. A submarine-hunting prototype was built in five days for a British visit, after which Johnson worked 72 hours straight and an order for 200 aircraft followed.
The technical range is wide. The U-2 won by flying above the reach of fighters and missiles rather than by out-fighting them. The SR-71 Blackbird required inventing fuels, sealants, paints, wiring and manufacturing methods, and it was designed by only 135 engineers. Less than 20% of its thrust came from the engine itself, which David uses to illustrate systems thinking. Cost gets equal attention: production engineers reviewed every prototype drawing, saving $10,000 to $20,000 per airplane on a 2,500-aircraft program, and a monthly cost review kept the books current. Johnson also held that designers should fly what they build, and said he needed to be scared once a year.
The contested material comes last. Johnson argued that design by committee yields nothing very stupid and nothing brilliant, and that the atomic bomb, Sidewinder and nuclear submarine came from unconventional methods. He also thought contractors should run their own bids, because commercial procedures often beat military ones. His 1985 forecast of fighter pilots flying unmanned aircraft from the ground is included. David closes by drawing parallels to Elon Musk and SpaceX.
Kelly Johnson's Skunk Works succeeded because he treated organizational design as inseparable from technical design: small empowered teams, minimal bureaucracy, direct authority, and speed as a design requirement. The host argues these principles still apply to modern engineering companies like SpaceX.
- Johnson's team built the XP-80 in 143 days with a small group, beating a 180-day schedule
- Johnson argued committees and paperwork erode the strong single-point authority that makes fast engineering possible
- Johnson's 14 rules and 20 host-derived principles recur across his career and parallel Elon Musk's SpaceX operating methods
Small, lean, empowered teams with one clear priority outperform large, overhead-heavy organizations in building aircraft.
- A breakthrough program is an organization before it is a design; Johnson's most important invention was the Skunk Works organization itself, built on small empowered teams, streamlined processes, and a culture willing to attempt untried things.
“A breakthrough program is an organization before it's a design.”
- Simplicity is a weapon; Johnson's KISS principle was a survival method for complex programs, and he repeatedly stressed minimizing paperwork.
“Make simplicity a weapon. He repeats this over and over again. His principle was KISS—keep it simple, stupid.”
- Extreme performance often comes from one brutally clear priority.
“Number 13, I think, is self-explanatory: Extreme performance often comes from one brutally clear priority.”
- Use a few exceptional people rather than many average ones; Johnson described Skunk Works as a few great people using straightforward methods with minimum overhead.
“Use a few exceptional people, not many average ones.”
- Program control should sit close to the person responsible for technical success, with managers given both authority and responsibility and minimal interference, so the builder can actually deliver.
“In his autobiography, he says he lets managers run programs with minimum interference while giving them both authority and responsibility.”
- The Skunk Works manager must be delegated complete control of the program in all aspects and should report to a division president or higher.
“The Skunk Works manager must be delegated complete control of his program in all aspects. He should report to a division president or higher.”
- Because only a few people will be used in engineering, ways must be provided to reward good performance by pay rather than by the number of personnel supervised.
“ways must be provided to reward good performance by pay, not based on the number of personnel supervised.”
- Design by committee, reviews, and consultations by those not directly doing the job produce nothing very stupid but nothing brilliant either, and major advances come from brilliant concepts developed by other methods.
“Nothing very stupid will result, but nothing brilliant either. And it's in the brilliant concept that a major advance is achieved.”
- The best companies are systems where everything affects everything else; on the SR-71, the engine inlet, airframe, and heat system formed one machine, and less than 20% of total thrust came from the engine itself.
“Less than 20% of the SR-71's total thrust came from the engine itself. The rest came from the inlet, the movable spike, afterburner flow path, and the cell design.”
- The first Skunk Works shop was built from engine boxes and a tent, and with about 120 people and 23 engineers, the XP-80 jet fighter was built in 143 days.
“The first Skunk Works shop was constructed of engine boxes and a tent. From this, the very first project, the XP-80 jet fighter, was built with just 120 people in 143 days. There were only 23 engineers on that project.”
- A subcontractor on one program had 1,206 people in quality control and 40 people inspecting, coordinating, and reporting for an instrument built by only 35 people, meaning overhead exceeded production labor.
“he discovered that there were 1,206 people employed just in quality control. This subcontractor had 40 people inspecting, coordinating, and reporting. They only had 35 people building the actual instrument.”
- Johnson's standard schedule was about 10-hour days, six days a week, with long hours discouraged by rule, which the host considers less demanding than Elon Musk's SpaceX schedule.
“With rare exceptions, long hours are not encouraged. From what I gather from the book, his standard schedule was about 10-hour days, 6 days a week.”
- Lockheed built a prototype submarine-hunting aircraft in five days for a British visit, made changes over a weekend, and Johnson worked 72 hours straight, which led to an order for 200 aircraft, the largest US aircraft production order to that date.
“The result: the Air Ministry gave Lockheed an order to build 200 airplanes. This was the largest aircraft production order placed up to that time in the United States.”
- There is no one good way to build all airplanes; the goal is to learn how to do things quickly and cheaply and to tailor systems to the degree of risk, and Tobi Lütke made the same point about multiple good approaches.
“There is no one good way to build all airplanes.”
- Skunk Works' purpose was to get results cheaper, sooner, and better through the application of common sense to tough problems; if it works, don't fix it.
“Our aim is to get results cheaper, sooner, and better through application of common sense to tough problems. If it works, don't fix it.”
- With small groups of good people you can work quickly and keep close control over every aspect of a project, and people challenged to perform at their best will do so.
“With small groups of good people, you can work quickly and keep close control over every aspect of the project. People challenged to perform at their best will do so.”
- The U-2 program employed 50 people across experimental and production engineering, and the SR-71 had only 135 engineers, a small team for such a difficult aircraft.
“On the enormously difficult SR-71, the one I've mentioned to you over and over again so far, there were only 135 engineers.”
- The contractor had 1,206 people employed just in quality control, while the SR-71 was designed with only 135 engineers.
“he discovered that there were 1,206 people employed just in quality control.”
- The Skunk Works operated on a 10-hour day, 6-day week with a scoreboard counting contract days; they delivered the XP-80 on day 143 against a 180-day schedule with 23 engineers on the project.
“On day 143, the plane was accepted by the Army Air Corps and ready for flight. We had beaten the schedule of 180 days.”
Single-person authority and organizational structure must stay minimal, since committees and companies suppress breakthroughs.
- Johnson's 14 Points were written to spread the basic operating rules of Skunk Works through his organization and to anyone it interfaced with, covering delegation, small offices, restricted access, simple drawings, minimal reporting, cost reviews, and pay for performance.
“As a way to spread these ideas throughout his organization and anybody they interfaced with, he wrote this thing called 14 Points.”
- Strong single-person authority essential to the operation is being eroded by committee and conference control from within and without, and immediate decisions require one person in charge.
“I see that strong authority that is absolutely essential to the kind of operation slowly being eroded by committee and conference control from within and without.”
- Johnson insisted on being able to call a counterpart who stays with the project to conclusion and get a same-day decision, which he said is not possible under standard operating procedure.
“He can call them up and get a decision that same day. Kelly Johnson says that's just not possible in standard operating procedure.”
- In the early days, when a failure was located and fixed, the team flew the next day with no committee inspection or review; the absence of process was the process.
“Well, we had located our trouble and we were ready to fly. No committee inspections, no review. In those days, when we were ready to fly, we flew. Next day, first flight.”
- During a Lockheed customer test flight, Howard Hughes seized the throttles and controls and stalled the airplane to zero airspeed, and Johnson ordered the pilot to take the controls back, establishing who ran the program.
“Hughes turned and looked at me as though I had stabbed him. I repeated, Milo, take this thing home.”
- When NACA refused Johnson access to its wind tunnel for speeds above 300 mph, Johnson went over their heads to General Arnold, who ordered the test run, saying to hell with the tunnel and to call him if it blew up.
“Put that airplane in the tunnel and run the test for Kelly. Find out what's wrong with my airplane. To hell with the tunnel. If it blows up, call me.”
- Major breakthroughs such as the atomic bomb, Sidewinder missile, and nuclear-powered submarine were achieved by methods other than conventional business, which committee design tends to suppress.
“Development of some of this country's most spectacular projects—the atomic bomb, the Sidewinder missile, the nuclear-powered submarine—all were accomplished by methods other than the conventional way of doing business.”
- For Kelly Johnson, technical design and institutional design were not separate; he treated the organization's structure as part of the engineering problem.
“he didn't separate technical design from institutional design”
- Most companies, while desiring the benefits, will not pay the price in revised methods and procedures for a skunkworks-type operation, and will not delegate authority to one individual.
“Most companies, while desiring the benefits, will not pay the price in revised methods and procedures for setting up a skunkworks-type operation.”
Speed is decisive, so bureaucracy should be pared away and work should begin before paperwork is perfect.
- Speed is a decisive competitive advantage; the SR-71 Blackbird, with almost no defense except acceleration, was never shot down and over 3,000 missiles fired at it missed, so the power of speed is underappreciated as a competitive factor.
“Over 3,000 missiles were shot at the SR-71 Blackbird and none hit. All it did was go faster.”
- Bureaucracy should be treated as an engineering variable; reporting, approvals, meetings, and customer visits are drag that must be pared away if they consume time without advancing the project.
“Bureaucracy is an engineering variable. Johnson treated reporting, approvals, meetings, and customer visits as drag.”
- Begin building before the paperwork is perfect; Lockheed's formal contract arrived 4 months after work began, and Skunk Works often started on a handshake when the need was urgent.
“Lockheed says the formal contract arrived 4 months after work had already begun. Skunk Works often began on a handshake when the need was urgent.”
- Secrecy can accelerate work by reducing outside interference and letting a compact group move fast; the Skunk Works team was warned not to discuss the project outside its small organization.
“Secrecy can accelerate work, not just hide it. Secrecy reduced outside interference and let a compact group move fast.”
- Speed is not a side effect but a design requirement, and it applies to project completion as well as to aircraft performance; Johnson's team delivered the P-80 prototype in 143 days.
“Speed is not a side effect. It is a design requirement.”
Redundancy, simplicity, and testing of consequences are required for safe frontier aircraft.
- Howard Hughes was an owner and avid pilot who was reckless, reportedly crashing planes three or four times, and once plowed through a thunderstorm on approach to Denver without warning passengers.
“On a flight as he was approaching Denver, Hughes encountered a big thunderstorm.”
- Decision makers who doubt a need should test the consequences themselves; Lieutenant Kelsey, who doubted the dive flap, was flown into a dive where compressibility left him unable to reach the switch, and after the tail broke off he bailed out, broke a leg, and became a believer.
“He put the plane in a dive and soon encountered compressibility effects so extreme that he could not reach the dive flap switch.”
- The host argues that modern air travel is the safest form of travel, while many people died or were injured as designers learned to make safe airplanes.
“How safe is air travel now? It's the safest form of travel for any humans, right?”
- Everything built must be able to relight, restart, and keep flying, because accidents at the frontier were frequent and redundancy became Johnson's obsession.
“With everything we build, we make sure that we can relight, restart, and keep flying.”
- The more complexity in a design, the more potential for problems, so Johnson reverted to KISS during the SR-71 design.
“I reverted to my old axiom: keep it simple, stupid. The more complexity, the more potential for problems.”
- Johnson served as flight test engineer alongside test pilot Eddie Allen, learning what to record and how to stay calm through dive tests, stalls, and spins on new aircraft.
“It was an excellent indoctrination to the art, skill, science, adventure, all that goes into flight testing.”
Problems must surface early through candid reporting, data-based challenges, and direct daily questioning.
- Junior engineers should challenge analysis with data; Johnson, early at Lockheed, challenged the stability analysis of an aircraft, went back to the wind tunnel, and solved the stability problem himself.
“Johnson challenged the stability analysis for this plane that they're building. He went back to the wind tunnel and then solved the stability problem himself.”
- A typical day began at 7:00 AM with a short, informal meeting with the engineers on key problems, asking what the biggest problem is and who is responsible for it.
“A typical day for me would start at 7:00 AM and I would immediately hold a meeting with the engineers working on key problems.”
- Mistakes are acceptable but concealed mistakes are not; mistakes must be reported immediately with recommendations for correction.
“Mistakes are acceptable, concealed mistakes are not.”
- Program management must include a monthly cost review that covers spending, commitments, and projected costs to program completion, avoiding books running 90 days late and sudden customer overruns.
“There must be a monthly cost review covering not only what has been spent and committed, but also projected costs to the conclusion of the program. Don't have the books 90 days late and don't surprise the customer with sudden overruns.”
- Johnson found the answer to an unstable Lockheed design after 72 wind tunnel runs, after first telling senior designers the airplane was unstable in all directions.
“It took 72 tunnel runs before I found the answer to the problem.”
Engineers who fly and build the aircraft themselves feel the risk and fix problems faster.
- Designers must ride with the pilot and feel the risk, because when their own life is on the line through a plane's testing they do better work.
“You have to ride with the pilot. When your life is literally on the line, you're going to do a better job.”
- Those who design aircraft should also fly them, since the engineer knows where marginal parts and flap behavior are, and sharing the pilot's and customer's concern gives a proper viewpoint; Johnson said he needed to be scared once a year to keep that balance.
“I've shared the concern of the pilot. I've shared the concern of the customer. I figured I needed to have the hell scared out of me once a year”
- When top leaders work alongside employees, including doing hands-on work, employees enjoy it and are more engaged, which is the parallel between Johnson and Elon Musk.
“The employees actually enjoyed the fact that their top boss cared enough to work right along with them.”
- Putting people in close contact with the aircraft as it is built lets them follow it through development and feel responsible for their parts, so problems get fixed quickly.
“It is our practice to put the people in close contact with the airplane while it's being built so that they can follow it throughout its development. They feel responsible for the parts they make.”
Frontier progress requires changing the envelope and inventing the methods needed to build the product.
- Do not optimize the old fight; change the envelope. The U-2 won not by dogfighting but by flying so high that fighters and missiles of the time could not reach it.
“The U-2 did not win by dogfighting, it won by flying so high that fighters and missiles of the time could not reach it.”
- At the frontier, you must invent the methods to make the product as well as the product itself; for the SR-71, that meant inventing special fuels, materials, sealants, paints, wiring, plugs, and manufacturing techniques.
“Whenever you're at the very forefront, you don't just have to invent the product, you have to invent the methods to create the product.”
- Stealth began as an operational idea rather than a technical one; Johnson had been thinking for decades about radar, visibility, speed, altitude, and exposure time.
“stealth began as an operational idea first. What they mean by that is he was already obsessed for many decades with thinking about radar, visibility, speed, altitude, and exposure time.”
Lifelong self-education, passion, and continuous open-minded learning shaped Johnson's engineering success.
- Johnson's childhood poverty, with nine children and work from age seven, combined with access to a Carnegie-funded library in Ishpeming, Michigan, led him to read widely and decide at 12 to become an aircraft designer.
“He returned an even richer resource. I went to the library almost every day. It opened a whole new world to me.”
- Johnson declined three offers of a Lockheed company presidency, saying leading advanced development projects was the best job in the corporation, because he had wanted to do this work since age 12.
“Three times I was offered a company presidency at Lockheed, and three times I declined it. To me, there is no better job within the corporation than head of advanced development projects”
- To pursue aeronautical engineering one had to study civil, chemical, electrical, and mechanical engineering first, a curriculum Johnson considered excellent because it gave a basic education in everything needed to design and build an airplane.
“It was an excellent curriculum because it provided a very good basic education in everything it took to design and build an airplane.”
- Johnson credited Tom Swift fiction and Carnegie library books on aircraft with setting his career goal, and read every aviation book he could find, spilling his enthusiasm over into his work.
“It became my goal to be just like Tom.”
- An engineer working at the frontier must keep up with advancing technology and constantly learn; Johnson reworked textbook problems on vacations and attended Caltech seminars for many years.
“It was and is important for an engineer to keep up with the advancing technology.”
- Johnson argued back against professors when he disagreed, which helped him establish respect and a good reputation when he later joined Lockheed.
“I was not so deferential that I would not argue back if I disagreed. And I did.”
- Keep an open mind and do not automatically write off any idea, because you never know which idea will work; Johnson's professor repeated this advice and Johnson used it his whole life.
“Do not automatically write anything off, anything.”
- Johnson's love of learning, shown in his reworking of textbook problems on vacations, kept him from losing mathematical capability and fueled his engineering edge.
“I was determined not to lose my capability in mathematics, and I enjoyed both those vacations.”
- Johnson grew up poor, worked from age seven, and was self-supporting from age 12, paying $7 weekly for room and board, which shaped a determined work ethic.
“By the time he's 12 years old, he's contributing $7 every week for room and board.”
- Andrew Carnegie's donated libraries in Ishpeming, Michigan, where his iron ore came from, opened a new world to Johnson and were the most important early influence on his life after his father.
“Next to my father, I credit Andrew Carnegie with being the most important influence on my early life.”
- Continuous learning drove Johnson's work; he remained obsessed with radar, visibility, speed, altitude, and exposure time for decades, which shaped his designs.
“the importance of this continuous learning that Kelly talks about over and over again”
Small concentrated customer offices and contractors with delegated authority and shared rules improve requirements and reduce misunderstanding.
- The customer must also be small and empowered; clean interfaces with the military or intelligence customer reduce misunderstanding and correspondence, rather than merely requiring lean contractor teams.
“Johnson's model did not merely demand lean contractor teams; it demanded clean interfaces with the military or intelligence customer. This reduced misunderstanding and correspondence.”
- The contractor can create the requirements; with the F-104, the idea came from Lockheed to the Air Force, and Lockheed submitted the proposal before the Air Force had an operational requirement for that kind of aircraft.
“the idea came from Lockheed to the Air Force, not the Air Force to Lockheed. Lockheed submitted this proposal before the Air Force even had an operational requirement”
- The customer should have a small, highly concentrated project office as a counterpart; the Skunk Works total for both the CIA and Air Force representatives on the U-2 and SR-71 was 6 people, versus 145 Army personnel on another program.
“There was another program where, just for the interface, there were 145 Army personnel just to interface with the designer and manufacturer. In contrast, the total at Skunk Works for both the CIA and the Air Force representatives for the U-2 and the SR-71 was 6 people.”
- The contractor must be authorized to test the final product in flight and should test in initial stages too; without this, the contractor rapidly loses competency to design other vehicles.
“If he doesn't, he rapidly loses his competency to design other vehicles.”
- The contractor should be delegated authority to obtain good vendor bids for subcontract work, because commercial bid procedures are very often better than military ones.
“Commercial bid procedures are very often better than military ones.”
- What the Skunk Works does is secret, but how it does it is not; Johnson wrote that he had long tried to convince others to adopt its principles, yet very seldom has the formula been followed.
“What the Skunk Works does is secret. How it does it is not. I've been trying to convince others to use our principles and practices for years.”
- Suppliers and others associated with a project must be given the same rules and permissions as the main program, which cuts red tape and cost and lets participants focus on the product rather than the system.
“Suppliers and others associated with the project must be extended the same kind of rules and permissions that are given us for the entire program.”
Leading people through respect and shared purpose works better than driving them.
- Johnson's promise to employees was a challenging, worthwhile job with stable employment, fair pay, advancement, a contribution to national defense, good management, sound projects, good equipment, and good work areas.
“I owe you a challenging, worthwhile job providing stable employment, fair pay, a chance to advance, and an opportunity to contribute to our nation's first line of defense.”
- Johnson's late-1930s philosophy of living had five parts: belief in God as sustenance in crisis, health as necessary for happiness, purpose in doing what you want well, a loving and understanding spouse, and respect from and for the people you work with.
“Respect of the people for whom you work and who work for you. That sums up my philosophy of living.”
- Lead people rather than drive them; drive yourself if you must, but not others, a lesson Johnson's mentor taught him after he early in his career failed to consider others' reactions.
“My mentor taught me that it's much better to lead people, not to drive them. Drive yourself if you must, but not anybody else.”
- To get the best performance from old-time machinists who want to discover methods themselves, put them in the experimental shop under the engineer's direction and make them party to developing the data.
“So in Skunk Works, we put them in the experimental shop under the engineer's direction and made them a party to developing the data.”
- Johnson's first wife was diagnosed with cancer young and attempted suicide with pills during depression, and he came to believe life should be shared to be meaningful.
“Much as I love my work and always have, perhaps more than most people, I always believe life should be shared to be really meaningful.”
Production alternatives and cost must be weighed in aircraft design decisions.
- Each engineering drawing released for the prototype was also sent to production engineers instructed to find every alternative way of making the part that did not hurt drag, maintenance, or cost, and three days were spent choosing the best production method.
“Every time we release an engineering drawing to our manufacturing director to build a part for our experimental airplane, we also release it to a group of production engineers”
- Johnson noticed the university wind tunnel sat idle most of the time, asked to rent it when not needed, and earned significant money renting it to car companies.
“He starts renting out the wind tunnel to car companies at the time and making a bunch of money.”
- Cost must be considered in aircraft design, since aircraft are becoming so expensive they are hardly worth what they do; Johnson's production engineering review saved $10,000 to $20,000 per airplane on a 2,500-aircraft program.
“Cost must be considered. Aircraft are getting to be so expensive they are hardly worth it for what they can do.”
Integrity requires returning money, abandoning impractical designs, and refusing work one does not believe in.
- Integrity is a performance multiplier; Johnson returned government money or declined work when he thought the result would not be effective, and said he never designed a plane he did not believe in.
“Johnson sometimes returned government money or declined work when he thought the result would not be effective.”
- Know when to kill your own brilliant idea; Johnson scrapped and canceled a hydrogen-powered plane once deeper design work showed it was too impractical and unwieldy.
“Kelly Johnson had this idea for a hydrogen-powered plane, but as they got deeper down the design path, he realized it was too impractical and too unwieldy, so he scrapped the design and canceled the project.”
- Johnson predicted in 1985 that the fighter pilot would eventually fly an unmanned fighter from the ground, saving aircraft cost and manpower and improving pilot safety, as fighter prices had exceeded $30 million per plane.
“I can foresee the day when the fighter pilot will be on the ground flying an unmanned fighter with a missile in it.”
Stress, secrecy, and the need for escape marked the personal cost of the work.
- During the U-2 program, after the CIA asked Johnson to publicly provoke the Soviets and he denied a shot-down aircraft was a U-2, security advised him to vary his routes and he kept a pistol at his bedside for years.
“It was about that time it was suggested to me by security people that I not go to work and come home by the same route”
- Johnson loved ranching, riding horses, and raising cattle, which he described as a life-saving escape from the pressures of work.
“It was life-saving escape from the pressures of work.”
- Johnson developed recurring stomach ulcers under work stress, and a fatal accident would bring one on within about 24 hours; one ulcer episode led to surgery removing half his stomach.
“With more responsibility came more ulcers. Whenever one of our aircraft had an accident, particularly a fatal accident, I would develop a stomach ulcer in about 24 hours.”
- Health is necessary for happiness: without it one cannot be truly happy, and Johnson's first wife's cancer diagnosis showed what happens when health is taken away.
“Health. Without it, one cannot be truly happy.”