Elon Musk's decision-making isn't instinct or luck. It's a filter: find the one constraint actually stopping you, run a fixed five-step algorithm against it, then check the result constantly instead of once. Most leaders skip step one and wonder why nothing sticks. The order is the whole trick.
"...I think I've said that phrase about 200 times."
When Elon Musk repeats himself, you listen. Even discounted, since he tends to round up, it's a phrase he leans on hard enough that it's worth knowing what it is.
This is what Elon Musk's decision-making comes down to: find what's actually stopping you, and fix that first.
Your stalled project tells the same story. You've seen it: the dragging deadline, circling meetings, the underperformer nobody wants to name.
Most leaders push harder. Elon finds the choke point first. You're not running a two-hundred-billion-dollar rocket company, and the logic still scales down to whatever is stuck on your desk.
Elon hunts the bottleneck — the single limiting factor jamming everything up. It's the missing key to the door right in front of you. Once identified, he unlocks it fast.
What Is Elon Musk's Limiting Factor?
That phrase is "limiting factor." Checking the transcript from the Cheeky Pint podcast, Musk uses it twenty-five times in one conversation. Not the two hundred he claimed afterwards, but enough to mean it.
Business theorist Eliyahu Goldratt named the idea behind it decades before Musk did. His Theory of Constraints holds that any system's output is capped by a single limiting factor, and improving anything else first is wasted effort, however good it looks on paper.
Musk just applies it faster than most.
Take the computing chips that power his AI ambitions. He sees two constraints stacked on top of each other. There aren't enough chips, and there isn't enough power to run the chips that do exist.
Tesla builds its own chips instead. Its space data centres skip the grid entirely and tap solar power around the clock. Between them, that sidesteps the factory backlog and the weather.
He's got this wrong before, and corrected it just as fast. In 2017, Tesla promised 5,000 Model 3s a week. It made 260 that quarter.
Musk had bet on a fully automated factory: robots doing the work, no bottleneck in sight. The bet was wrong. The real constraint was hiding in the battery assembly line, where a supplier's automation had simply failed to work.
Once he found it, Tesla took the line in-house and rewrote the software in what Musk called "20 to 30 man-years of software in 4 weeks." Production caught up soon after.
Finding the wrong constraint costs you a year. Finding the right one costs you a month.
That's the limiting factor test, in practice. Name the one thing actually capping progress, and solve for that before touching anything else.
How Do You Find Your Own Limiting Factor?
I cast my mind back to a repair-capacity problem I once had to unravel. The leaders before me had wrongly assumed the network had enough capacity to handle demand. No one had ever questioned it.
That assumption was the real limiting factor, hiding in plain sight.
Marc Randolph and Reed Hastings ran a smaller, faster version of the same test. Before launching Netflix, they wondered whether a disc would even survive the post. DVDs weren't for sale yet, so Randolph mailed Hastings a CD instead, in a plain envelope, to see if it would arrive in one piece.
When it did, the limiting factor disappeared, and the business launched.
Next time your own project stalls, run the same test. Name the one constraint, out loud, specifically. Then find the cheapest, fastest way to check if it's really the thing in your way. Only reach for a bigger fix once you know.
What Is Elon Musk's 5-Step Algorithm?
Musk laid this out to Tim Dodd, host of Everyday Astronaut, while explaining why a rocket's grid fins didn't fold in. He calls it an algorithm, not a framework — deliberately.
A framework describes a way of thinking. An algorithm executes, the same way, every time.
The five steps run in order, and the order is the whole point.
- Question every requirement. Know who set it, and question it — regardless of how senior or convincing they are. Only what survives the questioning gets to stay.
- Delete what you can. If you don't end up restoring at least 10% of what you cut, you didn't cut enough.
- Simplify what's left. Only once the deleting is done. Never simplify something that shouldn't exist in the first place.
- Speed it up. Cycle time only gets accelerated once the first three steps are finished — never before.
- Automate last. Automation locks in whatever remains, so it comes after everything else has been questioned, cut and simplified — never before.
Step one is where most leaders skip straight past. In one business I worked in, a repair job needed manual authorisation past a cost threshold. There were two thresholds; the lower one was meant as a checkpoint on parts use.
The call centre just handed out an approval code for it. No operator ever checked the job or questioned the parts.
Why would they? They weren't engineers.
Once that requirement got questioned, the flaw was obvious. We scrapped the lower threshold, replaced it with random audits by an in-house engineer, and call volume dropped by 50% from that one change alone.
Nobody had asked who actually needed that rule. Musk's own reason for putting this step first borrows from outside the algorithm entirely: strip a problem back to what's actually, fundamentally true, then rebuild from there, rather than trusting how it's always been done. That's reasoning from first principles, and it's the thinking Musk runs before he ever gets to step two.
Which Step Matters Most for a Leader Starting Out?
Musk runs this on the factory floor himself, not just in the five steps on paper. Isaacson describes him stopping mid-meeting to ask an engineer why a part needed four bolts — three might do. The answer he got back was "that's the way it's always been done," and it wasn't good enough for him.
That's the whole move. Not a redesign, not a budget request, not a project plan. One question, aimed at one requirement, asked out loud in the room where it gets decided.
Pick one recurring meeting or approval on your calendar this week. Find out who actually requires it, and why. Most of the time, the honest answer is nobody currently does. It just hasn't been asked in years.
Steal Elon Musk's Filter for Your Next Decision
One Good Decision gives you the tool to run it, step by step, this week.
How Does Elon Musk Use a Feedback Loop to Course-Correct?
"I'm a huge believer in taking feedback," Musk has said. "I'm trying to create a mental model that's accurate, and if I have a wrong view on something, I'll say — I used to think this one thing that turned out to be wrong. Now thank goodness I don't have that wrong belief."
That's not humility. It's mechanism.
A feedback loop only has two settings. It tells you the decision is working, or it tells you it isn't. There's no third result that leaves you exactly where you started.
Most of us only get that signal once, right at the end, after we've fully committed. The verdict arrives all at once, and it's either brilliantly right or painfully wrong.
Musk builds the check in earlier, and runs it constantly instead of once.
Every decision he makes is a hypothesis, not a conclusion. Tim Urban, who spent time watching him work, described the difference as a cook following a recipe versus a chef building one from scratch: one repeats what's proven, the other tests and adjusts as it goes, dish by dish.
None of Musk's decisions are meant to be final, then. Each one is a working answer, held loosely enough to be wrong, and swapped out the moment the feedback says so.
That only works if he expects to be wrong on a regular basis, not occasionally.
How Do You Build a Feedback Loop Into Your Next Decision?
Musk has put a number on it. In Walter Isaacson's biography, he reckoned he made close to a hundred calls a day, walking the factory floor.
"At least twenty percent are going to be wrong, and we're going to alter them later," he's said. "But if I don't make decisions, we die."
Twenty percent wrong isn't a failure rate. It's the cost of deciding at all, built in ahead of time.
Most decisions never get that number attached. They just get made, then judged much later, once it's too late to adjust cheaply.
Set the moment you'll check your next call before you make it — a week out, a specific meeting, a number that tells you plainly whether it's working. Decide first. Then decide when you'll decide again.
That second decision is the one most leaders skip.
What Order Does Elon Musk's Decision-Making Process Follow?
He's already told you he makes close to a hundred of these decisions a day. At that pace, there's no time to run five steps and a feedback check on the wrong problem.
That's why the order here isn't interchangeable.
Run the algorithm first, before naming the limiting factor, and you risk questioning, deleting and simplifying a process that was never the real constraint. All five steps executed well, aimed at the wrong target.
Run a feedback loop before you've actually decided anything, and there's nothing to check. Feedback only tells you whether a fix worked. It can't tell you whether you fixed the right problem in the first place.
So the sequence holds: find the one thing actually stopping you, before you touch anything else. Then run the algorithm on that thing specifically, not on whatever's easiest to fix. Only once a decision exists does feedback have something to correct.
Skip a step, or run them out of order, and you're back to spending effort (Musk's, or yours) on problems that were never the bottleneck to begin with.
How Do You Run This on the Decision You've Been Avoiding?
Go back to your stalled project.
Start with the filter, not the fix. Name the one thing actually stopping it: not the three or four things you'd list if someone asked, the single one that's actually jamming everything else. If you can't name it in a sentence, you haven't found it yet.
Once you have it, run Musk's five steps against that one thing — question it, delete what you can, then simplify whatever's left. Only then does speed matter, and automation comes last of all. It's one move, run in sequence, not three separate techniques. It's the same order Musk runs it in, and skipping ahead is usually why a fix doesn't stick.
Then set your check before you commit — a date, a meeting, a number that tells you whether it worked. Some of it will be wrong. That's already priced in, not a surprise to deal with later.
What changes here isn't the size of the decision. It's the order you attack it in.
The dragging deadline, the circling meetings, the underperformer nobody wants to name. Pick one, this week. Run the filter first. Then the algorithm. Then the feedback check. That's decision-making, worked in order.
What's the one thing actually stopping the decision you've been avoiding — and have you tested it, or just assumed it?
FAQs
Is Elon Musk's decision-making just first principles thinking?
No. First principles thinking is the lens, not the system. It tells Musk what's actually true; the limiting factor test, the five-step algorithm, and constant feedback are what he does with that clarity. Treat first principles as the starting point, not the whole method, or you'll copy his thinking without his results.
How do I know if I'm solving the wrong problem?
If the same issue keeps resurfacing after you "fix" it, you likely treated a symptom, not the limiting factor. Musk's test: name the single constraint out loud, specifically, before touching anything else. If you can't name it in one sentence, you haven't found it, and the next fix won't hold either.
What is Elon Musk's decision-making style?
His style looks impulsive: fast, blunt, occasionally reckless. The system underneath isn't. Style is what an outside observer sees; the system is the sequence, the constraint, and the constant correction running underneath it. Copy the system. The style is just how it looks from the outside, and it isn't the part worth borrowing.
What's the first step to applying Elon Musk's decision-making system?
Don't start with the algorithm. Start by naming the single thing actually stopping your stalled decision, in one sentence, before touching anything else. Once it's named, the five-step algorithm gives you somewhere to run it, and building the habit of moving on it fast matters more than getting each step perfect.
How does Elon Musk's decision-making differ from a typical corporate framework?
Most corporate frameworks describe how to think, then leave the applying to you. Musk's is an algorithm: the same five steps, run the same way, every time, with no room for interpretation. If you want the fuller contrast between frameworks and executable process, see how that plays out generally.
Steal Elon Musk's Filter for Your Next Decision
One Good Decision gives you the tool to run it, step by step, this week.
Written by
