10 Lessons

“…it’s important to know what the post-skinned cat looks like.”

Elon Musk

Walter Isaacson’s biography of Elon Musk is a fascinating read. It reminded me of his biography of Steve Jobs, an exciting personality who drives progress but is ultimately flawed. The book contains as many cautionary tales as advice for creative productivity.

I’m not one to follow the day’s fashion and pile into Musk for his opinions; most who do this engage in performance art, which I am allergic to. What is helpful is to look at some of the critical lessons one can draw from this book. Much of this is well-trodden territory, but the fact that it comes from the person behind some of the most stupendously successful enterprises in modern history gives them more weight.

These are by no means the only lessons to be drawn from the book, but these are, in my view, the most useful for a leader.

Idiot Index

Musk developed a quick heuristic to identify when a component had the opportunity for cost reduction. He called it the idiot index. It was an elegantly simple way to identify potential areas for improvement.


The bigger the delta, the more opportunity to simplify the manufacturing process. This is a powerful way to identify potential for improvement. It seems obvious, but the cost assumptions are built-in in many organizations, so few ask questions. The assumption is things are just expensive.

What about non-finished goods processes, though?  A similar approach might be total cycle time vs. task time; the lower the percentage of actual work required to complete the job vs. the cycle time indicates an opportunity for process optimization, or more likely removing non-value-added processing in the form of unnecessary bureaucracy and specialization.

Take approvals as well; what percentage of transactions are rejected? The lower that number, the bigger the question about the need for the approval gate is. Even outside of manufacturing, there are huge opportunities to question the efficacy of a process based on a simple delta.

Don’t do dumb things at scale

“We’re going to be doing dumb things, but let’s just not do dumb things on a large scale,” he told Cantrell. Instead of launching large payloads, as Lockheed and Boeing did, Musk would create a less expensive rocket for the smaller satellites that were being made possible by advances in microprocessors. He focused on one key metric: what it cost to get each pound of payload into orbit. That goal of maximizing boost for the buck would guide his obsession with increasing the thrust of the engines, reducing the mass of the rockets, and making them reusable.

In large organizations with many levels of hierarchy and specialization, in other words, “more referees than doers” as Musk puts it, there is a tendency to build large-scale plans, assuming growth and a need for scale, before a concept or process has been tested in the real world.

An adaptable organization is comfortable creating a small-scale test in the real world to understand how the idea will interact in the physical world. This requires a mindset of hypothesis and test, relying on the data and not just on whose brilliant idea it was.

All requirements should be treated as recommendations

The only rules are the ones dictated by the laws of physics. Everything else is a recommendation.

Commonly, requirements, standards, and rules accrete over time in organizations. Sometimes, these requirements are so distant from when they were invented that no one in the organization knows where they came from or why. I experienced this many years ago in a comical fashion. Having taken over a PMO group and survived my first few months, I was in a meeting with my peers. A discussion about some process or report that still needs to be completed came up.  Frankly, I don’t recall the report; it was some silly bureaucratic thing (Office Space fans should have TPS reports in mind here). My peers thought the PMO should handle this, but it was the first time I’d heard of this report. It had yet to come up in any project or be mentioned by any of the 40 project managers in my organization.  I probed a little bit about this report, why we needed to do it, and what the real need was. Was it missing for this one project, or was there a more systemic issue? I quickly learned that every project out of the 120 in flight had not submitted this report. I asked the group when this report was last created. They mentioned the name of someone I didn’t know. When pressed, they admitted that this person left the organization seven years ago! I asked about a statute of limitations on these rules, but my interlocutor didn’t see the humor.

In addition to the normal accretion of requirements and processes from within, there are externally derived requirements, such as regulations and legal requirements. Often, these real requirements need to be better understood by the organization. Often, the requirement or regulation becomes the default excuse. “It’s a legal requirement” has become the most used phrase. Now, there are a few things to consider here.  First, these so-called regulations or legal requirements often aren’t requirements. Second, the specific wording of the requirements will leave a wide range of possibilities to satisfy them. The first step should always be to interrogate the actual text of the requirement. You might find it can’t be found (problem solved!). The rest of the time, you can understand what is textually required and figure out the best way to satisfy it.

Avoid processes that throw stuff over the wall

To be the chief engineer for the Model S, Musk hired Peter Rawlinson, a genteel Englishman who had worked on car bodies for Lotus and Land Rover. Together they came up with a way to do more than merely place the battery pack under the floor of the car. They engineered it so that the pack became an element of the car’s structure. It was an example of Musk’s policy that the designers sketching the shape of the car should work hand in glove with the engineers who were determining how the car would be built. “At other places I worked,” von Holzhausen says, “there was this throw-it-over-the-fence mentality, where a designer would have an idea and then send it to an engineer, who sat in a different building or in a different country.” Musk put the engineers and designers in the same room. “The vision was that we would create designers who thought like engineers and engineers who thought like designers,” von Holzhausen says.

Most companies’ general change direction is to increase complexity via specialization. This means optimizing each process component (local optimization) rather than optimizing for the whole and, as a result, the necessary customer outcome.

One natural result of this increased specialization is organizational silos that result in a “throw it over the wall” mentality, where one group will push a product or output to the next group, knowing (either implicitly or explicitly) that it doesn’t meet the quality necessary for the finished product. This seems silly on its face, but from a bureaucratic perspective, this can have benefits by moving work off the plate of one group and putting it in a dead zone where everyone can reasonably blame someone else.

Any process that allows groups to throw stuff over the wall should, therefore, be avoided. Even if (and perhaps because…) the groups in question are diligent and work well together, this will be a natural consequence of the structure. It’s also worth pointing out another reason many organizations build in complexity. Cost reduction via outsourcing. There may very well be good reasons to outsource, but generally, cost reduction isn’t a good reason. It bakes in this silo mentality, and the very cost reductions so touted will incentives the outsource partner to be very litigious because their margins are so thin.

Make decisions or die

Throughout the spring and early summer of 2018, he prowled the factory floor, like he had in Nevada, making decisions on the fly. “Elon was going completely apeshit, marching from station to station,” says Juncosa. Musk calculated that on a good day he made a hundred command decisions as he walked the floor. “At least twenty percent are going to be wrong, and we’re going to alter them later,” he said. “But if I don’t make decisions, we die.”

So much of the modern corporate structure is dedicated to so-called data analysis. This can be useless as thousands of hours spent developing reports or PowerPoint decks or analysis of data into pretty, but ultimately useless, charts and graphs. Whole departments were created to work in a new tool (Tableau, anyone?). All of this work, while it may seem useful, drags decision-making. Musk understands that being active and actively making decisions is critical. There is risk in making decisions. However, a leader must be willing to be wrong, and those who work for him must have enough respect to continue to be led after the occasional failure.

Jeff Bezos’s concept of type 1 and 2 decisions is useful here. Type 1 decisions are those that you can reverse easily enough, so there is little need to marinate over the decision too much; make a call and see what happens. Type 2 decisions can’t be reversed, could have a big impact (positive or negative), and would require more thoughtful consideration. Now, Musk has avoided the issues with Type 2 decisions by being super engaged at the work level. This can have the effect of moving what would be Type 2 decisions into Type 1 territory, increasing the speed at which decisions can be made.

Always go to the floor

The switch to stainless steel allowed SpaceX to hire builders without the specialized expertise needed for fabricating carbon fiber. At its engine test site in McGregor, Texas, it contracted with a company that erected stainless steel water towers. Musk told Riley to reach out to them for help. One question was how thick the Starship’s walls should be. Musk talked to some of the workers—those actually doing the welding rather than the company’s executives—and asked what they thought was safe. “One of Elon’s rules is ‘Go as close to the source as possible for information,’ ” Riley says. The line workers said they thought the tank walls could get as thin as 4.8 millimeters. “What about four?” Musk asked.

One of the most common and most damaging paradigms in corporate America is the lack of leadership participation at the point of work. By not working on the floor, so to speak, managers and senior leaders then require more analysis and meetings, which suck up the time available, making it difficult to get to the floor. Because of the initial decision not to spend a great deal of time at the point of work, and as a general guideline, I use 50%; all other decisions that will cause a drag on decision-making are made. Perhaps worse, leaders are making decisions based on flawed data, and they need more connection with the workers to understand the impact of their changes. Rinse, repeat, and you have a recipe for massive dysfunction.

Musk understands this at an almost genetic level. He learns the business and the science, and, most critically, spends a huge amount of his time on the floor where he has a unique insight, credibility with the workers, and the safety to make the quick call and be wrong and survive.

The people who do the work are in charge of the work

The biggest change Musk wrought was to put the design engineers in charge of production like he had done for a while at Tesla. “I created separate design and production groups a long time ago, and that was a bullshit mistake,”

Recently, on the social media site X, a user posted a chart showing the growth of those in the medical field in the United States. It shows two types of roles: doctors and administrators. The growth in doctors has a small, steady trend upward. The administrators, however, have grown enormously, excepting the doctors in the system by order of magnitude. This is common to the medical field; [looking at education](https://www.thecollegefix.com/vanderbilt-has-1-administrator-for-every-2-students-analysis/?utm_source=substack&utm_medium=email), you see a similar graph: massive growth in areas not doing the work that serves the very purpose of the institution.

One implication of this, aside from massively increased costs for very little value, is that those administrators will find something to do and that usually involves managing those doing the actual work. This means we are increasing distance from leadership and expertise. When it’s critical, you must have the people doing the work accountable and in charge of the work and how it gets done.

When stuck, design the future

The 1337 engine One method Musk used when a problem got hairy was to turn his attention to designing a future version of the product. That’s what he did with Raptor a few weeks after McKenzie took over. He declared that they were going to shift their attention to making a whole new engine. It was going to be different enough that he didn’t even want to name it after a breed of falcon, like Merlin or Kestrel. Instead he decided to use a meme from the world of coding and call it 1337, pronounced “LEET.” (The numerals kind of look like those letters.) The goal was to get an engine that would cost less than $ 1,000 per ton of thrust and thus be, he said, “the fundamental breakthrough needed to make life multiplanetary.” The point of leapfrogging to a new engine was to get everyone thinking boldly. “Our goal is the great adventure engine,” he said in a pep talk to the team. “Does it have a chance of success above zero? If so, put it in! If we find that there are changes we made that are too adventurous, then we will back up.” The guiding principle should be to make an engine that was lean. “There are a lot of ways to skin a cat,” he said. “But it’s important to know what the post-skinned cat looks like. The answer is muscly and gnarly.”

A common and useful way to get out of a funk intellectually is to walk and clear your mind. Getting into a new environment and letting your mind wander can pull you out of a loop. Designing the future is similar, but here, you keep the team challenged without the difficulty of having to solve a real problem. Since it’s the future and not a problem that will be removed today, the team can take more risks and be more out of the lane.  This also has the benefit of potentially solving something or voiding fodder for hateful future discussions, so it’s a good effort.

People watch the leader very carefully and act accordingly

Musk brought some toys, including a robot that could follow a person with its eyes and another that could break-dance, to one of the design reviews in mid-July. He believed that toys could offer lessons; a little model car had inspired him to make real cars using big casting presses, for example, and Legos helped him understand the importance of precision manufacturing. Optimus was standing in the middle of the workshop, supported by a gantry. It slowly walked around him and deposited a box it was carrying. Musk then took the joystick controller and guided Optimus to pick up the box and hand it to von Holzhausen. After Optimus finished, Musk gave a gentle push to its chest to see if it would fall over. The stabilizers worked; it stayed upright. Musk nodded appreciatively and shot some video of Optimus. “Whenever Elon pulls out his phone to take a video, you know that you’ve impressed him,” Lars Moravy says.

There are a few aspects of this lesson. Most fundamental is that leaders must operate from a clear set of rules and moral standards. Employees will spot hypocrisy very clearly. I have experienced this issue many times where leaders would say one thing, but their actions belied their real beliefs. In an organization that touted its commitment to employees continuously, the real sense of the leaders came through when you saw that employees lacked sufficient parking, but managers had separate parking areas with plenty of spaces near the door, so while employees had to often park in the IKEA parking lot across the street, the leaders had nice close parking next to the building. This spoke volumes about what the org valued (which is, in this case, power and hierarchy).

A more subtle takeaway from this lesson is that leaders must communicate along multiple frames. Musk, taking out his phone to record something, sent a powerful message to the team without the usually hysterical team rah, a quiet and powerful message of appreciation for a job well done.

Decline starts when you stop taking risks

“This is how civilizations decline. They quit taking risks. And when they quit taking risks, their arteries harden. Every year there are more referees and fewer doers.”

“More referees and fewer doers.” That is a good description of most large companies that have had time to build a large bureaucracy. Additional layers in the organization, additional titles, departments, and specializations will inherently reduce the appetite and aptitude for risk. Departments begin to value turf protection and avoid anything that smells of risk to their position (and the continued growth of the bureaucracy). Given the complex nature of most businesses, there’s never a shortage of reasons not to do something, so most things are either killed outright or put through an obstacle course of tar and like the mammoths of yore, die.

I’d recommend reading the book, but here are the top 10 lessons I drew from it. No matter what your view on Elon Musk is, there is no denying that he’s been successful in many competitive businesses and, in more than one case, built a large enterprise in space (no pun intended) that the “experts” thought impossible to succeed in.