Skunk Works by Ben Rich and Leo Janos is the story of the first skunk works operations that ran out of the Lockheed shop in California. Headed by Kelly Johnson and then Ben Rich, they were the ones responsible for the stealth fighter, U-2 spy plane, and Blackbird. Beyond lessons in government bureaucracy and ground breaking engineering, the book gives a blueprint for how to make anything.
1/ Decentralized command. The Skunk Works shop was run by Kelly Johnson, who had two ways of doing things, his way or the highway.
One part of his style was to let people do their own work and do it well. The first rule Johnson explained to new recruits (and you had to be recruited) was:
“The Skunk Works program manager must be delegated practically complete control of his program in all aspects. He should have the authority to make quick decisions regarding technical, financial, or operational matters.”
Hire good people and get out of their way.
This was easier said than done when building aircraft for the Air Force, but Johnson and his successor Rich both did it.
Decentralized command is a consistent piece of leadership advice from Jocko Willink too. You need to empower people to lead and accomplish goals on their own. Phil Knight wrote about his early team at Nike, “I let them be, let them do.”
Most people miss this.
“One reason, I think, is that most other companies don’t really understand the concept or its scope and limitations, while many others are loath to grant the freedom and independence from management control that really are necessary ingredients for running a successful Skunk Works enterprise.”
When Louis C.K. made Horace and Pete he didn’t do a lot of rehearsing. They did table reads Monday, rehearsals Tuesday, and shot Wednesday and Thursday. It was different. It was fast. One friend who helped said that the actors even felt odd because they often went home early. This worked because decentralized command.
“Pressure is good for really good actors,” Louis said. He gave good people the right conditions. Even for Alan Alda, who Louis didn’t want. Then he decided, why not? “I love every single thing he’s done and he wants this, so I said, let’s just do it.”
It was Louis’s show. It was his idea. It was his money, and “you always want to help a little,” he said. But Louis never kibitzed. “He (Alda) invented that fucking character, it’s not what I had in my head, it’s something a billion times better.” Brian Koppelman said the same thing about his show Billions. Though he and David Levien wrote the show, it’s the director each episode that makes it something more. It’s decentralized command.
2/ Hit your MITs, and stop at good enough. Satisfy your most important things (MITs) – and stop. The Blackbird MITs were a speed of Mach 3 and altitude of 80,000 feet. It put the plane out of reach of Soviet missiles and Migs. MITs will evolve over time, and that makes getting them right a challenge.
Skunk Works managed this. The stealth bomber “was put together with avionics right off the aviation version of the Kmart shelf.” Ditto for the Blackbird, “to save cost and avoid delays, whenever possible we would use engines, avionics, and flight controls from other aircraft and cleverly modify them to fit ours.”
The MITs for a prototype were: cheap, reliable, repeatable.
They used proven equipment that was lying around. As the plane evolved, the designers, engineers, and machinists adjusted their focus to performance and quality. The titanium skin, for example, forced a retooling of the shop.
Don’t worship a MIT. Just get there and move on to the next one. Someone suggested they extend the 3,000 mile range of the plane another 80 miles. Johnson did the calculations and noted that those extra miles would cost $20M and the plan was scrapped. Rich wrote:
“If we were off in our calculations by a pound or a degree, it didn’t particularly concern us. We aimed to achieve a Chevrolet’s functional reliability rather than a Mercedes’s supposed perfection. Eighty percent efficiency would get the job done, so why strain resources and bust deadlines to achieve that extra 20 percent, which could cost as much as 50 percent more in overtime and delays and have little real impact on the overall performance of the aircraft itself.”
Rich pointed out that some things were paramount, but going past them wasn’t.
In episode 026 of my podcat we looked at the early strategies of Warren Buffett, who shifted from one MIT to another. Initially Buffett wanted cigar butt companies. His MIT was intrinsic value which was greater than the market cap. Later the MIT was the quality of the business and management. He realized that certain gains in these MITs were good enough. He wrote in an early letter:
“It does also not seem sensible to me to trade known pleasant personal relationships with high grade people at a decent rate of return, for possible irrigation, aggravation, or worse at potentially higher returns.”
For the Blackbird the MIT was mach 3 and 80,000 feet. Once they got there they stopped.
For Buffett the MIT was “a decent rate of return.” Once he got there he stopped.
3/ People. When you are dealing with people (which is always) remember that people are human and will do dumb/silly/crazy/unexpected things.
For as great as the Blackbird was, it still had a problem, the pilot. They would return from missions and debrief that they “ran out of ass before they ran out of gas,” and “my mind went numb ten minutes ahead of my ass.”
The plane had incredible technology, but people were involved and people make mistakes.
When Jack Schwager was asked why his book Market Wizards holds up he said that even though the market changes, the people don’t. Charlie Munger wrote, “what determines the behavior are incentives for the decision maker,” and “getting the incentives right is a very, very important lesson.”
“Begin each day by telling yourself: Today I shall be meeting with interference, ingratitude, insolence, disloyalty, ill-will, and selfishness – all of them due to the offenders’ ignorance of what is good or evil.” – Marcus Aurelius
People can be amazing too. The Blackbird made a sonic boom as it flew, and was kept away from populated areas. Except one, Susanville California. It just couldn’t easily be avoided. One day the townspeople (sans-pitchforks) had enough. They complained about plaster cracking and windows shattering from the bursts.
Rather than preach their authority, the Skunk Works team “had the townspeople in, showed them the airplane, appealed to their patriotism, and told them the boom was ‘the sound of freedom.’ They lapped it up.”
4/ Prioritize and execute.
How do you eat an elephant? One bite at a time.
How do you build a spy plane? One piece at a time.
A frequent implementation of P&E is the beat the shit out of it and fix what breaks system. This is what Yvon Chouinard did at Patagonia. “We test until something fails, strengthen that part, then see what fails next.” It’s also what they did at Skunk Works.
In 1961 they were testing how much pressure Blackbird’s fuel tanks could hold. They did this by pumping air into the titanium tanks. Like dad at a birthday party, they wanted to see how much air could fit into their titanium balloon.
Of course they had to do this at night “when very few people were around.” They started to pump and the crew hid behind a thick steel wall. At ten inches of mercury pressure “Kaboom! The drag chute compartment in the rear blew out.” They fixed the fail point and started again.
At ten and a half inches “the drag chute forward bulkhead ruptured with a loud bang.” Eventually they got the design they needed. Breaking is a good prioritize and execute technique.
Labeling things as too hard is another. “If something is too hard,” explained Charlie Munger,” we move on to something else. What could be simpler than that?”
A third way to P&E is to dissect big problems. Stephen Dubner explained why smaller problems are easier to solve.
- They may be virgin territory.
- They are less tangled.
- They are easier to change.
- They are more accurately observed.
We saw this with Nike and Phil Knight.
- Running shoes were virgin territory, “to go out for a three-mile run was something weirdos did.”
- Knight and his co-founder Bill Bowerman were only making one kind of shoe.
- To launch a shoe company was easier than a shoe + apparel or shoe + apparel + technology.
- The market shortcomings for running shoes was clear to Knight and Bowerman, there weren’t many shoes and those that were out there weren’t good.
Thanks for reading, I’m @mikedariano on Twitter. If you don’t like reading posts, you could listen to my notes intead; https://soundcloud.com/mikesnotes, https://itunes.apple.com/us/podcast/mikes-notes/id1055386383, https://overcast.fm/itunes1055386383/mikes-notes.