Elon Musk’s 5-Step Algorithm (Most People Run It Backwards)

Elon Musk has a five-step algorithm for solving problems, and most of us run it backwards, which is why we end up with processes that are bloated, automations that don't work very well or are overly complex, and half of the stuff that we're building probably shouldn't exist in the first place. And I'm going to share the algorithm itself, but more importantly, I'm going to share why the order matters more than the steps themselves.

Here's why I pay attention to Elon in the first place. I am fascinated by people who repeatedly do things that are next to impossible, right? Like, very few people have built a $100 million company, fewer have built a billion-dollar company, and almost none of us have built multiple billion-dollar companies across different industries, like Tesla, SpaceX, Starlink, now xAI. And I would personally argue he's, you know, one of the greatest entrepreneurs ever, certainly of our generation.

When someone pulls off something improbable once, it's like, you can chalk it up to luck, right? It's like somebody asking, "How do you find the perfect spouse?" Well, if you've done that once, that's fine, it doesn't mean that process is repeatable, and if you've done that multiple times, then chances are you're not very good at explaining how to find the perfect spouse, right? But when someone does something very effective that's like, improbable to begin with, and they've done it more than once, it's a track record. You know, you look at it and you go, okay, at some point, you got to acknowledge it's not fucking luck, right? And that's worth studying to me.

So I've dug into how Elon makes decisions, and I've, you know, almost obsessively, and I've found what he calls "The Algorithm," and it's basically five steps that he developed the hard way as an actual engineer through solving the wrong problems at SpaceX and Tesla before figuring out the right approach. And here's the process:

Question every requirement. Look at the thing, look at the process, question every requirement that's there, especially requirements that come from really smart people because smart people have got a track record of being right, and that lets them make assumptions that nobody challenges. And then those assumptions become invisible requirements that are baked into whatever it is that you're building, and you've got to surface them.

Delete any part of the process that you can. Subtract first. Our natural tendency is to add stuff. "Okay, well we'll just bolt this on to make it easier, and then we'll add this to change that, and then you add this and then you add that," and before you know it, you've got more complexity, you've got all these different integration points, you've got room for error, and changing anything gets more complex. So delete whatever you can get away with deleting. In fact, there's a rule that he has: if you're not adding back 10 percent, you're probably not deleting too much.

Simplify and optimize. Only after you've questioned everything and then deleted do you ask, "Okay, how do I simplify what's left?" Because if you try to optimize or simplify first, you're potentially engineering something that didn't need to exist in the first place.

Accelerate cycle time. Now that you've got the right requirements, only the steps that you need, and the simplest version, how do you run it faster? How do you get the same output in less time?

Automate. Automate last.

Success is in the sequence here. I cannot tell you how many times I've seen people automate something—I've done this myself, recently, on a process—where they skip the first four steps and work the process backwards. In fact, Elon says he's worked it backwards, and that's how he discovered this process: you look at something and you go, "Hey, that seems like a huge pain in the ass. You know what we should do? We should start automating this, and we should start, you know, adding some AI to that, and we should make that process more efficient, and we should try to simplify, and then okay, now let's try to delete some steps from there."

Now that there's all this complexity and automation built into it, deleting actually gets more difficult, and so you've got these big, complicated machines, and that's what you're trying to simplify. That is the result of working this process backwards. That's the sin.

If you're tech-friendly, if you have an engineer mindset, if you love systems thinking, you're probably doing this too. And the real genius to me of this algorithm isn't just the five steps; it is the order. Because if you flip the order, you're going to get a very different outcome. It's question first, then delete, then simplify, then accelerate, then automate absolutely last.

I've been hammering this with my team recently—my ops person who runs our automations and all of that stuff—and we've actually rolled this into part of our values as part of an EOS implementation because I believe the tendency is, especially for tech-minded and systems-minded people, to love the process of building so much that we lose track of the actual output—the velocity and quality of that output. We fall in love with the machine and the process instead of the output. If you take these five steps and run them in order across any part of your business—content, sales, customer experience, internal ops, automation—it will have a massive impact on you. So I hope it helps. Adios.