The Five Step 'Algorithm' Driving Tesla’s Success

0:03A lot is written about Elon Musk. What did he say when you told him you were going to write a book about your experience? I got a thumbs up emoji. Thumbs up emoji. That's like an endorsement. This week on Bold Names, we're talking to Jon McNeill. He's the former president of Tesla. He's got a new book out called The Algorithm, and it's all about the lessons he learned from working closely with Elon Musk. That's next. From the Wall Street Journal, I'm Tim Higgins. This is Bold Names, where you'll hear from the leaders of the bold name companies featured in the Wall Street Journal. Today we ask, what's the secret sauce to the Elon Musk School of Management?

0:46John McNeil, welcome to Bold Names. Thanks, Tim. Excited to have you here today. You've had a lot of big jobs, but you're probably best known for the time you served as president of Tesla when the company brought out some really big vehicles. The Model X SUV, you worked to bring out the Model 3. This was a game-changing vehicle. And I'm just curious, what did you learn about management from your time working so closely with Elon Musk? Man, how much time do you have? This is one of those four-hour podcasts. We're changing bold names. There's so much to unpack there. But so there were no management playbooks for this, no case studies.

1:29Like we literally were making it up as we went. And 10Xing a company at that scale in hardware is, as we all found out, super-duper hard. Because it's not just a matter of adding servers. We had to add factories and supply chains and delivery centers and service centers. Like the whole shoot and match had to be added. So we developed, through a series of mistakes, we developed this algorithm that we used inside that's basically the operating system of Tesla. And it weakly drives both problem solving and innovation so that we could just run at a pace that a company at that scale and in that industry had never run in before.

2:10Okay, I want to talk about the algorithm because that's what your new book is about. But first, let's just kind of get into your mindset as you joined the company, because you had not been a car executive, right? You had been an entrepreneur. Up until that point, you had started and sold something like six startups before joining Tesla. I'm curious, how was that experience different from what it was going to be like at Tesla? You're going into a company, you're working for Musk. How is that different? I've been a CEO through those six companies. So I hadn't like supported it in myself in a long time.

2:47And so I had to learn to be a subordinate. And that was really in support of a CEO. And so all the times that I was asking teams to do unreasonable things for me as a CEO, the payback was almost sudden and instant as I stepped into that job. And I think, you know, Musk hires orthogonally in the sense that he doesn't hire people from the industry. because he wants fresh thinking that is not weighed down by old constructs. The head of supply chain came from Apple. The head of battery cell innovation was homegrown. He'd come up through the ranks of Tesla having no experience in automotive before.

3:34The head of engineering came from Segway and Apple. So it was a bunch of people who were hired orthogonally, and I was sort of in that same bunch. and so I had to quickly learn manufacturing at that scale in the supply chain. But so the mentality I was in was basically the same mentality I was in as an entrepreneur, which is you wake up every day, you've got a stack of problems that are begging for attention and you figure out which problem is the biggest and you pull that off the stack for that day and try to solve it. Let's break down how you were dealing with those problems or those opportunities because that's really what the book is about, with this algorithm that kind of emerged from Musk's school of management is all about.

4:22Tell me what the algorithm is first. The algorithm is this process that we use to drive innovation and growth at Tesla. And it was really arrived at, it's a five-step problem-solving device that we use. And then I write about three secret ingredients. ingredients. What are those five steps? Yeah. So it is, first of all, question every requirement. Like when you're given a set of requirements, question the hell out of them. Are these like a requirement of law, of safety, of physics? And if you can't find a positive answer to those three questions and you really start to rip out these requirements because you make the problem more simple.

5:06And it's important to simplify the problem so you can actually scale it. The second step is to then radically simplify the process. And that is rip out every step that the customer doesn't pay you for. So in our world, the customer doesn't pay us for engineering change orders. They don't pay us for documentation. They don't pay us for filling out finance forms. They don't pay us for any of that. They pay us for the car. So you essentially rip out every step that the customer is not paying you for. That's second step. Third step is now that you've got a simplified process, and this is going to sound crazy from a tech company, but you trial that process manually.

5:45You do not automate. And the reason you don't automate is because if you don't have the process figured out, you're going to automate something and you're going to get to a bad answer more quickly than you would otherwise. Or you're going to take that bad process and essentially cement it because it's really hard to change automation after you start. Once you get that process down, then you speed it up, which shows all of its flaws. That's the fourth step is add speed. And then the fifth step is you add then automation last. Once you've got this thing operating at speed and then it will scale and be repeatable.

6:22I'm just curious, when was the first time you heard this term, the algorithm? Did Musk make you get a tattoo? How did this kind of filter its way out through the company? Our weekly management team meeting was essentially riffing on lessons that we've learned or problems that we were in the middle of. And then we would try to, over time, build frameworks for those. and uh and we did a lot of reflection on mistakes uh and and so this thing this five-step process really didn't have a name because we were we were kind of developing it and learning it uh and uh and then applying it ourselves and one day i forget who named it um but somebody said hey we ought to like call this something uh so that we can really communicate it to everybody to the troops, everybody involved.

7:14And somebody came up with, let's just call it the algorithm. Like, this is how things work here. And so it's stuck. So no tattoos. I don't have any permanent markings on my body, but I probably have permanent scar tissue from figuring this thing out by mistake. Just ahead, John explains why automation comes last. When we were experimenting on the factory floor, we would literally tell the teams, until we've really figured out how to optimize this process, do not bolt the machines to the floor. Because once you do, they get really hard to move. And how he learned this the hard way. Stay with us.

8:01You know, it's interesting, the first step, a question every requirement, really feels like it comes out of kind of first principles thinking. It does. And what is first principles thinking to you? So I'll give you an example. And this is really an example of how we learned the algorithm by mistake. So when we were trying to produce the Model 3, there was this separator between the battery and the compartment of the car, passenger compartment. And the separator, we could not figure out how to get this thing manufactured. It would delaminate, misshape, all sorts of stuff. We couldn't get it to lay flat across the battery.

8:40And Elon was like personally involved in this. So we're like in the manufacturing cell trying to fix this. And one day we look up and said like, who specced this thing? Like who said we needed this? This is like after three weeks of like sleeping in the factory trying to fix this. The people that were closest by were the battery guys and they said, oh, this comes from the vehicle dynamics team. They told us that we needed this for sound deadening. So we go to the vehicle dynamics team. We're like, why did you guys spec this for sound deadening? They said, we didn't. This comes from the battery team.

9:13They spec this for flame retardancy. So we go back to the battery guys. We're like, we were just with the vehicle team. They told us they didn't spec it. They said, you guys did. They said, we didn't spec it. We said, okay, what engineer's name is on the spec? So we find the engineer's name. Maybe that guy. Yeah, but this is part of the process. You question the requirement first. And then you say, who thought this was required? Because maybe they have a good reason. We should go talk to them. We go try to find this particular individual. We can't find them because they were a summer intern that wasn't even at the company anymore.

9:50Tim, we had spent three weeks of our lives in this manufacturing cell trying to fix something that wasn't necessary. and it led us after several of these like big incidents like this to say we got to start every problem solve with questioning the requirements and finding out who name of the person spec this so that we'd go talk to them and see if it's really real or not which then gets to the second delete every possible step yeah exactly you know in one hand it's like this pretty simple of an idea like get rid of the things you don't need but how do you optimize a company and its processes without cutting too close to the bone where everything falls apart, right?

10:29Well, I think you could argue we did that. We did that systematically because in that deletion process, we started to delete things that we couldn't identify that the customer would pay us for. And oftentimes, like Elon would say, hey, I think unless we have to add back 10 % of what we deleted, we probably haven't gone far enough. So I would come in and say, hey, we deleted X, Y, and Z out of this process. And he would come back and say, like, have you added anything back yet? And I'd say, no. And he'd say, okay, go back and see what else you can delete. And so oftentimes we would kind of get close to the bone, maybe into the bone.

11:05And then we'd pull back and say, okay, like we've gone too far. Let's have this step back in. Almost failure is kind of telling you how far you've gone. Totally. It's telling you where the guardrail is for sure. Yeah. Which I think gets to this idea of you don't want to automate until you've run it a few times to make sure it works, right? Exactly. because what automation does is it kind of puts cement around that current process because it's really hard to change software and automation after you've done it. And when we were experimenting on the factory floor, we would literally tell the teams, until we've really figured out how to optimize this process, do not bolt the machines to the floor because once you do, they get really hard to move.

11:42Well, and this was problematic with the rollout of the Model 3, right? Right. This was a huge issue. Right. This is where you learned some of these lessons, right? Totally. Or step five came into play. Why? What happened? Man, you nailed it. Automate last was the big learning from the first iteration of the Model 3 production. So Elon was talking about the alien dreadnought and the machine that makes the machine. And he had this goal of creating the most automated factory on the planet. And part of the reason is we had seen the factories in China and we'd seen how much more advanced they were than us and the rest of the Western manufacturers.

12:19We said, we got to fix this. And let's just take a step in here. Let's put a little bit more color around this period of time, right? The Model S, the Model X, these are the vehicles that Tesla had at that point. Really a niche carmaker at this point. An electric carmaker selling really high-end vehicles. The Model 3 was going to be a game changer. More importantly, it was going to save us from bankruptcy because we were out of cash. A bet the company product. A bet the company move. So what we do is we start Model 3 by saying, let's create the most automated factory in the world. And to do that, a digital twin was created, a SIM.

12:59And the entire factory was simulated digitally. Did they have a little simulated Elon running around? Yeah, they didn't, but that's a pretty good question. Yelling at people? That would have been funny to suggest, actually. But once the thing was built, you could see, oh my God, practical things have totally been overlooked. Like the machines were so close together that a human being couldn't get in between them to maintain them. Like the maintenance guys couldn't get in there. I remember touring it, and I felt like being in a submarine where it was like going through these narrow passages. Yeah.

13:31It was tight. Yeah, and because things broke all the time, you couldn't fix them. The cars were falling off the trays, onto the floor. It was multiple levels. It was a disaster. Yeah. And we could not produce a car on that line. In fact, that line ended up being scrapped. Uh, and, uh, and so Jerome Guillen, who was an executive at the company and charged trucks. And who later become a president. He took my role after I left and he, he stepped in and said, guys, we have to go back to manual process. Like we got to prove this manually. So I'm going to build a tent over the weekend outside and we're going to start to make manual threes by hand or model threes by hand.

14:12and that's where we learned this lesson for the 10th time, automate last. But this was like this Model 3 line was our biggest and most expensive mistake ever, almost killed the company. Like we were literally down to the wire of having nickels and dimes left in our bank account before we got those Model 3s flowing out of the factory. You know, it's interesting because if you talk to a traditional car guy, traditional car executive, they probably would have said, you know, automation can be challenging, right? And I have talked to people over the years, and they probably said to Elon, hey, this is a bad idea.

14:50But I think it gets to some of that corporate culture that you write about, that if you look at every no as a potential yes, that sometimes there's good reasons for that no. Yeah. And I think culture has a lot to do with it, for sure. We were a bunch of software people, essentially. And so we were hell-bent on automating everything because that's the answer as we come from our background. So we even carry some of that in, which led to things like I just described, where you've got now an over-automated factory because you've got a bunch of people that are super excited about automation doing it.

15:30And so I think the problem can come from the yes and the no in the sense that we were the yes to this technology. And a traditional car person may have been a no or a caution. And you can have both ends of that equation battling progress. Since you've left Tesla, you're now on the GM board. So you still have a seat, a front row seat at kind of how the industry is evolving. And since your days at Tesla, the electric car market has just changed a lot, right? It has been a little bit of a roller coaster. What's the current state of the EV market here in the US? This is one area of technology where the rest of the world is ahead of us, both in terms of their supplier, that companies being able to supply are like really unique.

16:13It's China, the world, and then the US, which is not letting China in at this point. Yeah, it's like China, and then I would say Europe and North America, probably on two different trajectories right now. And so when I was getting to know Mary Barra and Mark Royce at GM. Mary's the CEO of GM and Mark is the president. and extremely talented leaders. And the one question I had was, I wanted to figure out how seriously they took product. Because I've just learned over time, and this is a mantra in Silicon Valley that I think Steve Jobs helped to solidify, which is compelling product wins. In fact, if you have close to perfect product, you don't even need a sales force because the world beats a path to your door.

16:59And at a time, we had that at Tesla. We'd had zero advertising budget. And yet we're growing, you know, 10Xing. And so I wanted to see like how compelling the vehicles were in the product plan at GM. And so we went vehicle by vehicle. And I was like, oh, my gosh, these are super compelling EVs, like really compelling EVs. I got to drive some of them. And they had leapfrogged Tesla. and those cars have come onto the market and GM has gone from zero to now the second ranked EV manufacturer behind Tesla. Really compelling cars, like a$30 ,000 bull to$33 ,000 Equinox. They're really cool cars up to like a Cadillac Escalade, which has every bill and whistle you could imagine on a car.

17:48And so I think the situation we find ourselves in in the US is that we have administrations going from one end of the spectrum to the other, sort of a 180 on policy. EVs have become very political, right? Whether it's the Democratic administration trying to push everybody into EVs or the current Republican administration saying something different. Something different, exactly. And so you've got this base of users that I think grows over time, even in the U.S., even with policy that is negative towards EVs. And so we're certainly not going to have 50 % of the people driving EVs, I don't think, by 2030 or 2035.

18:27But it's going to happen over time because it's just a superior product. It's interesting. You have this very interesting experience in the automotive world. Yeah. Right. You have worked for two of the biggest names of this modern generation of cars, Mary and Elon. How do you compare them? I would say you have two very talented engineers.

18:53and in Elon, you've got a visionary that pursues super ambitious goals. And in Mary, I think you kind of have the same thing. You've got this visionary that said, I see the car market moving. We've got to get there. And so I'm going to move this 100-plus-year-old organization towards that. But I think what I appreciate about working with Mary is she is hands down the best leader I've been around. And her ability to, number one, attract talent is world-class, sort of similar to Elon. Her ability to get the most out of that talent, also world-class, so similar to Elon. But her ability to exercise kindness while she does it is very unique.

19:40and she's somebody that like almost everybody that meets her has a universal reaction, which is, I want to be on her team. I want to support her because she's so compelling and caring and you can tell she's genuine. And I so respect her leadership and think in a lot of ways, she and Elon are both world-class, but in a lot of ways, she's a little bit the antithesis of Elon, too. She's just a very compelling person in her own right. Coming up, John talks about the darker side to his time at Tesla and why he decided to walk away from the company. We had a Christmas holiday where I was finally getting a break and I was with my family and my family sat down and did sort of an intervention on me, which I was not expecting at the holiday.

20:32That's next.

20:48Okay, so Tesla runs on this operating system called the algorithm. Yeah. Does the algorithm work without an Elon Musk? It does. And that's why I wanted to write this book. And the book is told, each chapter of the book is told through the eyes of frontline employees who are deploying this framework daily without supervision. I got to say that that was one of the things that I found very interesting in this book is if you're a recruiter out there, you just give it a roadmap of some of the biggest talent of Tesla from the last generation, right? Oh, yeah. But it's like there are dozens and dozens and dozens of these people.

21:24But the reason I told the story this way is I wanted people to get the notion that you do not have to be Elon Musk to do this. You do not have to be a CEO to do this. You can do this if you're leading a small team. You can do this if you're a team lead of 30 folks on an assembly line. If you're running 20 folks in a delivery center. These stories were meant to say, hey, Elon gets a lot of that. It gets almost all the press and the accolades. But here's the people that are really doing incredible work and driving incredible breakthroughs. And those are the stories I wanted to tell. So people got that notion that you don't have to be Elon to do this.

22:05Walter Isaacson, who wrote the biography of Elon, he's telling more the story of Musk. But there's this idea of great successes, of great breakthroughs that Musk has had. On the flip side of that, there are these demons that he is struggling with. And the question is, does he do what he does? Is he successful because of this other thing that's kind of eating at him? I guess I'm wondering, from your perspective, one of the things that Musk had the tools in his toolbox was he could motivate people through fear, right? Yeah, yeah. Yeah, and what I've discovered is over time there are different leadership styles, and those leadership styles can be really effective.

22:50Like the fear leadership style can be effective. and uh but what i've also discovered over time is you're totally uh inauthentic a phony if you're trying to copy somebody else's leadership style so i can't pull off that leadership style so i had i had to lead with my own leadership style while working alongside elon uh which made us kind of frickin frack next to each other uh and as it was at times probably pretty entertaining for people maybe frustrating at other times, but he does motivate at times through fear. Not all the time, but at times through fear. You should clarify that. Yeah, exactly.

23:30But it's not always. Because people will tell stories of really being impressed with his ability to get into the details of really hard engineering issues and just being inspired by what he can do. Exactly. Or being in a meeting. I've never laughed so hard with another executive like I've laughed with Elon. Like he's so funny when he gets in a mode of really riffing on through one of these problems. And then his sense of humor kicks in. It's really fun to work with him. But at times he pulls on that fear lever. I don't have that play in my playbook. So I had to discover my own leadership style.

24:02And that's another message in this book is that there are a myriad of leadership styles at work. You got to run yours as a leader. You cannot run somebody else's and try to mimic it because it just is inauthentic. Because you ultimately, your time ended at Tesla because you write in the book, you were coming home angry. Your wife kind of said, hey, this is changing you. Yeah. How is it changing you? We had a Christmas holiday where I was finally getting a break and I was with my family and my family sat down and did sort of an intervention on me, which I was not expecting at the holiday. And they said, you're becoming angry and mean.

24:43And I got to say, still chokes me up to hear that because that's the exact opposite of how I want to be remembered as a human. I don't want anybody thinking about their time working with me or interacting with me as somebody who's angry and mean. And so I literally went back in the first week of January and sat down with Elon and said, look, here's what's happening to me as part of this role. It's time for me to go. And I'll give you months and months and months of transition, but here's why it's time for me to go. A lot is written about Elon Musk. What did he say when you told him you were going to write a book about your experience?

25:21I got a thumbs up emoji. A thumbs up emoji. That's like an endorsement. Totally. And it's largely because he and I had this kind of tongue in cheek challenge with each other. We love first principles, both of us. We love to be minimalists. And so we took this challenge on between the two of us to run the company with three sentence emails between the two of us. So were these run-on sentences that are like Jack Kerouac style? No. Eventually we got down to characters, like how many characters were you communicating? And so his thumbs-up emoji was definitely in the spirit of our three-sentence emails.

26:03I totally knew that was an endorsement for sure. Well, John, this has been great. Thumbs up and congratulations on the book. Thanks, Tim. We reached out to Elon Musk and Tesla for comment. They did not respond. And that's bold names for this week. Our producer is Danny Lewis. Our video producer is Alexis Moore. And our fact checker is Aparna Nathan. Michael LaValle is our sound designer. Jessica Fenton wrote our theme music. Our supervising producer is Katie Ferguson. Additional thanks to Jana Heron. Our development producer is Aisha Al-Muslim. Chris Zinsley is the deputy editor. For even more, check out my columns on WSJ.com.