Love him or hate him, Elon Musk has built companies like Tesla and SpaceX that introduced a radically different way of building cars and rockets. There's a story about Elon that I love. Early on, he flew to Russia hoping to buy a rocket. The engineers there laughed at him. On the flight home, he started breaking down the cost of a rocket, material by material. He realized that the physical components of a rocket cost only about 2% of the total cost. The rest, administrative costs, bureaucracy, and layers of inefficiency. That insight helped spark the idea that it eventually became SpaceX. Elon says that thinking behind companies like Tesla and SpaceX follows a formula he calls the algorithm. So what is that algorithm and what might happen if more of us started thinking that way? Hi everyone, I'm Lynne Toman and this is Three Takeaways. On Three Takeaways, I talk with some of the world's best thinkers, business leaders, writers, politicians, newsmakers, and scientists. Each episode ends with three key takeaways to help us understand the world and maybe even ourselves a little better. Today I'm excited to be joined by John McNeill. John has spent his career building and scaling companies. Before joining Tesla, he founded and sold six startups. So he knows firsthand what it takes to turn bold ideas into real businesses. At Tesla, he served as president working closely with Elon Musk. Today John is a venture investor and the author of the wonderful book The Algorithm, where he lays out the thinking behind Tesla and SpaceX. He also explores how that approach to solving problems can apply far beyond cars and rockets. John, it's great to have you on the show. Thank you for joining Three Takeaways today. Thanks for having me. It is my pleasure. John, Elon talks about his five step algorithm for building things. What is the first step of that algorithm and how did it shape the way Tesla was built? This algorithm got developed over time basically through the mistakes that we had made and we did a lot of time riffing and reflecting on mistakes that we've made. And the first step of the algorithm comes from a number of experiences and that is question every requirement. Ask if those requirements are a requirement of law, of physics, or safety. And ask for the name of the person who came up with the requirement. So you can go interrogate whether that is really true or not. We were riffing one day on digital sales and we had limited amount of money. We could only open so many stores. We'd opened several hundred around the world. And then we started to brainstorm, could we sell a hundred thousand dollar product online site unseen? How would we do that? And one of the things we always concentrated on was the friction that we would put between ourselves and the customer. And friction online is measured in clicks. And Elon asked me, how many clicks does it take to buy Tesla? And I happen to know, I said 64. And he said, just for fun, just for kicks, pull out the Domino's app. Let's figure out how many clicks it takes to buy a pizza. And it's about 10. He's like, let's get it down to 10. And I said, well, 44 of the 60 clicks are in one document and that is the loan release document. And it's because those loan documents are like dozens of pages long. But let me figure out if there's a way around that. And so I went and questioned the requirement of every paragraph that was in a loan document. We had a great member of our legal team go through this with me and for me. But he came back and he said, you wouldn't believe this. Almost the entire loan document is not a requirement of law or regulators. It's well-meaning corporate attorneys who are trying to protect their bank. But none of this stuff matters. So I went back to the next week's brainstorming session with Elon. I'm like, do you know what? What I just heard was 12 pages of loan docs that everybody assumes are required, aren't required. It can be done in one paragraph. So I went and talked to like 10 different banks. They all slammed the door in my face. And then we finally got to a bank in Minneapolis, US Bank. And they said, we'll do it. We'll do a one-click loan. And we got 44 clicks eliminated immediately because we questioned the loan doc. Like who would be crazy enough to question paragraphs and loan docs? So we were crazy enough to do this sort of thing. So that's first step in the algorithm. If you're going to have a breakthrough, it gets a lot easier if you remove requirements that aren't real. Can you share more of what that looked like in the design of Tesla cars and how Mattel toy cars with just a top and bottom piece became a kind of inspiration? Elon gave us this challenge. Could we take 50% of the cost out of building a car? So he wouldn't ask for 5%. He wouldn't ask for 10%. He'd ask for something ridiculous because it takes you to another level of thinking. So the way a factory is laid out is they're often more than a mile long and they're kind of long rectangles. And you can go to the halfway point, the 50 yard line of a car factory, and you can look to your right and you've got hundreds of robots building the skeleton of the car. It's called the body shop typically. And you can look to your left and you've got thousands of people hanging parts on that skeleton. And that's called General Assembly. Doug Field, who is head of engineering and I walked out on the scaffolding that sits above the floor so we could start the brainstorm about how we could take 50% of the cost of the car out. And we were doing this because we'd been to China and we'd seen how cost effective China was and quite frankly we were scared to death. And so this wasn't just like an exercise out of thin air. It was an exercise we believed in long term survival. So Doug and I are out there looking at the factory and we look to our right and we see all these robots building the skeleton of the car. And then Doug's like, I got an idea. So he comes back the next day and we're in a conference room and he rolls a matchbox car across the conference table and says, here's the idea. And I'm like, what a toy car? What is this? And he said, well, in the body shop where we're building the skeleton of the car, it's all robots welding about 300 parts together. That's not the way matchbox cars are built. They're casted. Somebody pours liquid metal into a mold and a car comes out. What if we could cast the cars? And I knew enough about the physics to say Doug, the reason you can't cast cars is kind of obvious. You can't pour molten metal the size of a car into a mold and have the mold not melt and have the pressure of that situation not like explode a factory. Like there's a reason why matchbox can do that because they're the size of our thumb, but you can't do that with bigger metal pieces. He's like, I know, but I think it can be solved. And so he turns to me and says, do you have anything we can melt? I'm like, yeah, I got a bunch of scratch and dent rims in the factory that we reject. And we put them out behind the factory and a recycler comes and picks them up every week. We can melt those. He's like, we're going to get a few engineers and we're going to start to melt these wheels. And then we're going to take that molten metal and we're going to pour it into small molds first, figure out how to do small molds. And then we'll figure out how to make those molds bigger and bigger and bigger. So eventually they figured out how to cast half of a car skeleton, but it's all because Doug had this like insight and then pursued the insight in small steps. And we failed and blew up stuff and it wasn't linear, but we eventually got there. So now when you look at a Tesla car factory, whether you go to Austin, Texas or Berlin or Shanghai, you see that there is no body shop anymore. Not the factors gone because there are no robots that are welding 200 or 300 parts together because two parts come together to make that car. And we couldn't have foreseen a second order benefit, but the first order benefit is you remove a ton of complexity and cost. The second order benefit was when you're welding 200 or 300 parts together, the skeleton never quite aligns. And so you have to really do a lot of work to get the doors that you're going to hang on that skeleton to align and the windows and the seals and all kinds of stuff. When you have two parts that are cast and you put them together, everything fits every time. And so all of a sudden the doors fit, the windows fit, the gaps are right. Doug literally changed car manufacturing and now eight years later, every car manufacturer in the world wants to get their hands on casting, but they can't. It's really hard to do. So Tesla has built this compounded advantage over time with that one challenge that take 50% of the cost of the car out. Not to mention many fewer repair issues and problems. Exactly. And all the manufacturing issues that come with those 300 pieces not aligning very well. Something else that stands out that Elon doesn't hire people who've worked in the auto industry or in the space industry. Why is that? Is it because they carry too many assumptions about how things have to be done? He hires orthogonally. So what he means by that is he hires people that might have some related insight, but it never really worked directly either on the problem or maybe in the industry before. And it's not 100%. There are exceptions to the rule for sure. But for the most part, nobody came from the industry. And the reason for that is he didn't want you coming in with a preconceived notion of how the industry worked, could work, what was possible, what wasn't possible. That turns out to be a hell of an advantage when you get a fresh set of eyes and or people that just don't know enough not to be crazy enough to consider other solutions. And Jim Farley at Ford just talked about this over the past couple of weeks. He described a tear down of a Tesla they did and they tore down their leading EV, the Mustang E. And he said all of a sudden when you start to tear down those vehicles next to each other, you start to realize, oh, these people had no car experience, therefore they weren't biased to do things that car people would do. And the example he gave is the nervous system of a car is called the wiring harness. And it's literally hundreds of pounds of wire that gets strung around the car. So different things work like your headlights and your music and your seat and the AC, etc. When they did the tear down, they realized that the Tesla wiring harness weighed 76 pounds less than the Ford wiring harness. That's a very big deal because that's basically half a human you have to carry around in the car. And so that really affects the car's range. And Farley said, I knew why it happened because car people at Ford never questioned pulling weight out of the wiring harness. So they would just call the supply chain people and say, I need a wiring harness and they would order one up. Whereas the Tesla people were like, no, like to get range out of the car, like we got to completely rethink the way this wiring harness works. And so they had completely redesigned the wiring harness to save weight. And Farley said, that is the example of why having non-car people involved makes sense because these people had come from building phones and laptops where the weight really matters. And so they thought about how to be super efficient with the electronics they were designing. And that's just one small example of how orthogonal thinkers can be really productive. John, you had weekly meetings with Elon every Tuesday. Did knowing you and the team had to report progress directly to him light a fire under the entire organization? What were those meetings actually like? So there are a couple of kinds of meetings when he comes to Tesla. And I do think this aspect that you're pointing to, Lynn, is the I think this is the thing that people will come to understand and probably write about decades from now in terms of what makes him such an effective leader of fast-moving companies. He picks like the one or two things that are existential to the company, and then he only works on those. And they become the focus of his time at the company. For example, right now at Tesla, that is autonomy and robotics. So he'll show up and he's only working on those two issues. And the teams that are working on those issues have to report weekly progress to him. And if you're a team meeting with the CEO, you do not bring your B game. You bring your A game. And if something's going south, he knows every week. It allows him to keep momentum up. And so Elon can allocate capital where it's needed super quickly because he's seeing it firsthand. It's not coming through reports or presentation decks. He's seeing it firsthand. One thing you've noted is that Elon can sometimes just sit there in silence during meetings. What's going on in those moments? He's basically processing the problem. He is almost like a computer. He's taking inputs, processing those inputs and really deeply thinking about them. And he's not worried about the awkwardness of silence in that situation like a lot of us would be. For him, he's got to just stop and be quiet to process. And he knows that about himself. And so you get used to these moments where he literally is just trying to devise the next step based on the inputs that he's just heard. Elon is famous for setting targets that seem wildly ambitious. Was that deliberate setting goals so big that the only way to get there was to question every underlying assumption? Yeah, exactly. You don't question every underlying assumption if you have to grow the business 2% or 3%. In our case, we were doubling the business every eight months. So we're going from 2 billion in sales. In 30 months, we're at 20 billion in sales. So we 10x it. So if you're asking teams to double every eight months, they can't think incrementally. They have to think quantum leap. And so the way you're setting goals actually then determines how your team is going to be thinking about achieving those goals. And although incrementalism is important, he really wanted quantum in the big, big levers of the business. One of Elon's rules is to simplify everything and delete every possible step in a process. How did that play out at Tesla when somebody buys a car? We tried to delete as many steps in that process of car buying as possible. Like anybody who's bought a car really doesn't say it's a lovely experience. They go back and say that was worse than going to the dentist because they had so many steps, so much paper to fill out, going back and forth on pricing, et cetera. So we literally deleted every step in that process we could. We eliminated haggling. We had one price for everybody, including us in the company. We didn't get a discount on the cars. We pay what you pay. We eliminated loan docs, lease docs because we got down to a single paragraph. We automated the whole licensing process when you showed up and got your car that was actually a license plate on it. We just innovated all the way through that process and basically lined up every step in the process. They're paying you for a car. So let's eliminate everything that is not building the car because that's the only thing we get paid for. And the rest of it is all administrative overhead and junk that the customer doesn't see, doesn't care about, isn't going to pay for. And when you do that, when you map your process and you circle the stuff the customer actually pays you for, turns out it's very few things in their mind they're paying you for. And that's the mentality that we use as we go to eliminate and delete everything that's not in the customer's economic equation. At one point, Tesla did something that sounded almost crazy. It built an entire assembly line run by humans instead of machines. Why did Tesla do that? Because we'd learned too many times when you automate first, it is almost impossible to get the process right. And we'd literally just been through this. We had tried to create an alien dreadnought, the machine that creates the machine, and we had built a factory digitally and designed all the machines digitally. And when that factory was built, that line was built. It didn't work. And we realized we had violated this principle that we weren't going to violate again and that is automate first. We said, you can only automate last. And by that, we mean you got to do things manually, kind of third step of the algorithm. You run a process manually and you get it really, really efficient. And then you speed it up manually to show all the works. You add speed and then finally you're to process the works. And so we did this in the Model 3. We had to abandon this automated line that didn't work. And the company needed the cash from Model 3 or else we weren't going to survive. So we had to quickly get Model 3s out the door. And so a tent was built in the factory parking lot. And an assembly line was built in the tent by hand. And we started to build these things by hand so we could figure out the most efficient way then to automate and then eventually build the automated line. And we started to produce 100 cars a week and then 400 and then 500 on our way to 5000. And it literally saved the company to do that. So step one, question requirement, step two, you delete steps. Step three, now we're going to run the process manually. We're going to get that as efficient as we can. Then we're going to speed it up with a lot of cycle time so we'll try to double through. And then lastly, you automate because if you automate first, that often doesn't work. So the algorithm was developed really in response to this mistake of over-automating the Model 3 factory. Stepping back after working with Elon as president of Tesla, what's the biggest lesson you took away about building companies? I think how unfair the advantage is that comes from world-class team members. He has a very high bar for hiring. And what I learned was not compromising on talent upfront means that you build a world-class team and that world-class team is like an unfair advantage because they can do so much, so fast at such a high quality level. The ability to, him to attract engineering talent and executive talent is a key to his success because he does attract just unbelievably world-class people around him. John, what are the three takeaways you'd like to leave the audience with today? I think number one, anybody can drive innovation and drive product breakthroughs and process breakthroughs by applying the algorithm. Takeaway number two is, I think it's really important for you to use your own product because you start to realize the holes in that product if you use it on a daily basis. And I'm surprised how many companies don't use their own products. And then takeaway number three is set very ambitious goals because you change the way people think about the problem. If you're asking them to make a quantum almost impossible change versus an incremental change. John, this has been wonderful. Thank you so much. I really enjoyed your book, The Algorithm. And I have to say, I also drive a Tesla, which I really appreciate. Oh, thank you. I'm glad. If you're enjoying the podcast and I really hope you are, please review us on Apple Podcasts or Spotify or wherever you get your podcasts. It really helps get the word out. If you're interested, you can also sign up for the Three Takeaways newsletter at threetakeaways.com where you can also listen to previous episodes. You can also follow us on LinkedIn, X, Instagram and Facebook. I'm Lynn Toman and this is Three Takeaways. Thanks for listening.
